| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <CANFwon0DRqfUFVw9YU52En6A-ri_XT0kSgOt6pfR-W6s-LuPOQ@mail.gmail.com>
Date: Wed, 14 Sep 2011 16:47:37 +0800
From: Hui Zhu <teawater@...il.com>
To: linux-kernel@...r.kernel.org
Cc: Marc Khouzam <marc.khouzam@...csson.com>,
Thiago Jung Bauermann <thiago.bauermann@...il.com>,
Steven <mqyoung@...il.com>, colyli@...il.com,
Christoph Hellwig <hch@...radead.org>,
Steven Rostedt <rostedt@...dmis.org>,
Randy Dunlap <rdunlap@...otime.net>
Subject: KGTP (Linux Kernel debugger and tracer) 20110912 release (quickstart available)
KGTP is a realtime and lightweight Linux Kernel GDB debugger and tracer.
It makes Linux Kernel supply a GDB remote debug interface. Then GDB in
current machine or remote machine can debug and trace Linux through
GDB tracepoint without stopping the Linux Kernel.
Now, it supports X86-32, X86-64, MIPS and ARM.
Now, KGTP 20110912 release.
You can get the package for it from
http://kgtp.googlecode.com/files/kgtp_20110912.tar.bz2
or
svn co https://kgtp.googlecode.com/svn/tags/20110912
The main change of this version is add quickstart for ubuntu and
fedora, you can access gtp_quickstart.txt or
https://code.google.com/p/kgtp/wiki/Quickstart
To get other change info please goto https://code.google.com/p/kgtp/wiki/UPDATE
According to the comments of Christoph. I make a patch for Linux
Kernel and make it looks OK with checkpatch.pl.
Thanks,
Hui
Signed-off-by: Hui Zhu <teawater@...il.com>
---
Documentation/trace/gtp.txt | 1440 ++++++
Documentation/trace/gtp_quickstart.txt | 250 +
arch/arm/include/asm/gtp.h | 13
arch/mips/include/asm/gtp.h | 18
arch/x86/include/asm/gtp.h | 18
include/linux/perf_event.h | 3
kernel/events/core.c | 16
lib/Kconfig.debug | 11
lib/Makefile | 2
lib/gtp.c | 7323 +++++++++++++++++++++++++++++++++
scripts/getgtprsp.pl | 102
11 files changed, 9194 insertions(+), 2 deletions(-)
--- /dev/null
+++ b/Documentation/trace/gtp.txt
@@ -0,0 +1,1440 @@
+ Linux Kernel GDB tracepoint module (KGTP)
+ =========================================
+ By Hui Zhu <teawater@...il.com>
+ https://code.google.com/p/kgtp/wiki/HOWTO
+ 2011-09-12
+
+Table of contents
+-----------------
+
+What is KGTP
+Report issues about KGTP
+Get info about GDB tracepoint
+Install GDB for KGTP
+Get KGTP through http
+Get KGTP through svn
+Config KGTP
+Compile KGTP
+Install KGTP
+Uninstall KGTP
+Use KGTP with DKMS
+How to get new version GDB
+Howto use
+ Exec it
+ Make GDB connect to gtp
+ If GDB on current machine
+ If GDB on remote machine
+ How to use /sys/kernel/debug/gtpframe
+ How to use /sys/kernel/debug/gtpframe_pipe
+ Get the frame info with cat
+ Get the frame info with getframe
+ Get the frame info with GDB
+ Add module symbols to GDB
+ How to use getmod.py
+ How to use getmod
+ Get GDB tracepoint commands
+ Get register info from Kernel
+ Get the value of variable from Kernel
+ How to use use tracepoint condition
+ How to use trace state variables
+ Simple trace state variables
+ Per_cpu trace state variables
+ Special trace state variables $current_task,
+ $current_thread_info, $cpu_id, $dump_stack, $printk_level,
+ $printk_format, $printk_tmp, $clock, $rdtsc, $hardirq_count,
+ $softirq_count and $irq_count
+ Special trace state variable $no_self_trace
+ Trace the function return with $kret
+ Use $ignore_error and $last_errno to ignore the error of tstart
+ Use $cooked_clock and $cooked_rdtsc the time without KGTP used
+ Use $xtime_sec and $xtime_nsec get the timespec
+ How to use performance counters
+ Show all the traced data of current frame
+ Get backtrace info(stack dump) from Kernel
+ Howto let tracepoint output value directly
+ Output stack dump directly
+ Switch collect to output the value directly
+ Use printf command in actions
+ Get status of KGTP from Kernel
+ Set the trace buffer into a circular buffer
+ Do not stop tracepoint when the GDB disconnects
+ Howto show a variable whose value has been optimized away
+ Linux kernel "Compile with almost no optimization" patch
+ Update your GCC
+ Offline debug
+
+
+
+
+What is KGTP
+------------
+
+KGTP is a realtime and lightweight Linux Kernel GDB debugger and tracer.
+
+It makes Linux Kernel supply a GDB remote debug interface. Then GDB in current
+machine or remote machine (see "Make GDB connect to gtp") can debug and trace
+Linux through GDB tracepoint without stopping the Linux Kernel.
+And even if the board doesn't have GDB on it and doesn't have interface for
+remote debug. It can debug the Linux Kernel using offline debug (See "Offline
+debug").
+It supports X86-32, X86-64, MIPS and ARM.
+
+For new user of KGTP, please go to see gtp_quickstart.txt.
+
+
+
+
+Report issues about KGTP
+------------------------
+You can post it in https://code.google.com/p/kgtp/issues/list or write Email
+to teawater@...il.com.
+
+
+
+
+Get info about GDB tracepoint
+-----------------------------
+Please goto http://sourceware.org/gdb/current/onlinedocs/gdb/Tracepoints.html
+
+
+
+
+Install GDB for KGTP
+------------------------
+The GDB that older than 7.3 have some bugs of tracepoint. And some functions
+of GDB are not very well.
+So please goto https://code.google.com/p/gdbt/ to install GDB for KGTP.
+
+
+
+
+Get KGTP through http
+---------------------
+Please goto http://code.google.com/p/kgtp/downloads/list OR UPDATE to download
+the package.
+
+
+
+
+Get KGTP through svn
+--------------------
+Some people have trouble with access to KGTP website. You can access kgtp
+through svn:
+
+------------------------------------------------------------
+svn checkout http://kgtp.googlecode.com/svn/ kgtp-read-only
+------------------------------------------------------------
+
+kgtp-read-only/tags/ Present for each release of KGTP.
+kgtp-read-only/trunk/ Present for the main trunk of KGTP.
+
+
+
+
+Config KGTP
+-----------
+
+Before compiling KGTP, you can choose which kernel you want build with and
+which compiler you want by making changes to the Makefile in your KGTP
+repository.
+For example:
+
+-------------------------------------------
+KERNELDIR := /lib/modules/`uname -r`/build
+CROSS_COMPILE :=
+-------------------------------------------
+
+KERELDIR is set to the directory which holds the kernel you want to build for.
+By default, it is set to the kernel that you are running.
+CROSS_COMPILE is set the compiler that you want to build the KGTP. Empty mean
+use current compiler.
+ARCH is the architecture.
+
+------------------------------------------
+KERNELDIR := /home/teawater/kernel/bamd64
+CROSS_COMPILE :=x86_64-glibc_std-
+ARCH := x86_64
+------------------------------------------
+
+KERNELDIR is set to /home/teawater/kernel/bamd64. Compiler will
+use x86_64-glibc_std-gcc.
+
+
+
+
+Compile KGTP
+------------
+
+For normal use:
+
+---------
+cd kgtp/
+make
+---------
+
+
+
+
+Compile KGTP with old Linux Kernel
+----------------------------------
+
+Most of time, KGTP can auto select right options to build with old
Linux Kernel.
+But if you want config special options with yourself, you can read following
+part.
+
+--------------------
+make AUTO=0
+--------------------
+With this option, KGTP will not auto select any build options.
+
+--------------------
+make AUTO=0 FRAME_SIMPLE=1
+--------------------
+With this option, KGTP will use simple frame instead of ring buffer.
+The simple frame doesn't support gtpframe_pipe. Use it can make KGTP can build
+with old Kernel that doesn't support ring buffer.
+
+--------------------
+make AUTO=0 CLOCK_CYCLE=1
+--------------------
+With this option, $clock will return rdtsc value instead of local_clock.
+
+--------------------
+make AUTO=0 USE_PROC=1
+--------------------
+With this option, KGTP will use procfs instead of debugfs.
+
+The options can use together, for example:
+-----------------------------------
+make AUTO=0 FRAME_SIMPLE=1 CLOCK_CYCLE=1
+-----------------------------------
+This build command make KGTP build success with Linux Kernel 2.6.18.
+
+
+
+
+Install KGTP
+------------
+
+------------------
+cd kgtp/
+sudo make install
+------------------
+
+
+
+
+Uninstall KGTP
+--------------
+
+--------------------
+cd kgtp/
+sudo make uninstall
+--------------------
+
+
+
+
+Use KGTP with DKMS
+------------------
+
+--------------------
+cd kgtp/
+sudo make dkms
+--------------------
+This commands will copy the files of KGTP to the directory that DKMS need.
+Then you can use DKMS commands to control KGTP.
+Please goto http://linux.dell.com/dkms/manpage.html to see how to use DKMS.
+
+
+
+
+How to get new version GDB
+--------------------------
+
+The old version GDB such as GDB 7.2 have some bugs about tracepoint that KGTP
+need. So I suggest you update your version older than 7.2.
+If you use UBUNTU, use can get the new version from
+https://lkml.org/lkml/2011/6/4/65
+
+
+
+
+Howto use
+---------
+
+Exec it
+-------
+
+If you have installed KGTP in your system, you can:
+
+------------------
+sudo modprobe gtp
+------------------
+
+Or you can use the kgtp module in the directory.
+
+-------------------
+cd kgtp/
+sudo insmod gtp.ko
+-------------------
+
+
+
+Make GDB connect to gtp
+-----------------------
+
+
+If GDB on current machine
+-------------------------
+
+---------------------------------
+sudo gdb ./vmlinux
+(gdb) target remote /sys/kernel/debug/gtp
+Remote debugging using /sys/kernel/debug/gtp
+0x0000000000000000 in ?? ()
+---------------------------------
+After that, you can begin to use GDB command trace the Linux Kernel.
+
+
+If GDB on remote machine
+------------------------
+
+---------------------------------------------
+#Open the KGTP interface in current machine.
+sudo su
+nc -l 1234 </sys/kernel/debug/gtp >/sys/kernel/debug/gtp
+(nc -l -p 1234 </sys/kernel/debug/gtp >/sys/kernel/debug/gtp for old version
+netcat.)
+#Let gdb connect to the port 1234
+gdb ./vmlinux
+(gdb) target remote xxx.xxx.xxx.xxx:1234
+---------------------------------------------
+After that, you can begin to use GDB command trace the Linux Kernel.
+
+
+
+How to use /sys/kernel/debug/gtpframe
+-------------------------------------
+
+This interface supplies trace frame in tfile format (GDB can parse it) when
+KGTP is stop.
+Please goto "Offline debug" to get more info about how to use it.
+
+
+
+How to use /sys/kernel/debug/gtpframe_pipe
+------------------------------------------
+
+This interface supplies same format trace frame with "gtpframe".
+But it can work when KGTP is running. And it is a consumer like "trace_pipe".
+
+
+Get the frame info with cat
+---------------------------
+
+sudo cat /sys/kernel/debug/gtpframe_pipe > g
+Then all the trace frame will be saved in file "g".
+
+
+Get the frame info with getframe
+--------------------------------
+
+KGTP package include a program "getframe" can help you save the trace frame to
+files.
+Following part is the help of it:
+
+-------------------------------------------------------------------
+getframe -h
+Get the trace frame of KGTP and save them in current
+directory with tfile format.
+Usage: ./getframe [option]
+
+ -g n Set the minimum free size limit to n G.
+ When free size of current disk is smaller than n G,
+ ./getframe will exit (-q) or wait some seconds (-w).
+ The default value of it is 2 G.
+
+ -q Quit when current disk is smaller than
+ minimum free size limit (-g).
+
+ -w n Wait n seconds when current disk is smaller
+ than minimum free size limit (-g).
+
+ -e n Set the entry number of each tfile to n.
+ The default value of it is 1000.
+
+ -h Display this information.
+-------------------------------------------------------------------
+
+
+Get the frame info with GDB
+---------------------------
+
+----------------------------------------------------
+#connect to the interface
+(gdb) target tfile /sys/kernel/debug/gtpframe_pipe
+#Get one trace frame entry
+(gdb) tfind 0
+Found trace frame 0, tracepoint 1
+#Get the next one
+(gdb) tfind
+Target failed to find requested trace frame.
+(gdb) tfind 0
+Found trace frame 0, tracepoint 1
+----------------------------------------------------
+This way is better to work with python to parse Kernel.
+I will introduce them in example.
+
+
+
+Add module symbols to GDB
+-------------------------
+
+Sometimes you need to add a Linux kernel module's symbols to GDB to debug it.
+Add symbols with hand is not very easy, so KGTP package include an GDB python
+script "getmod.py" and a program "getmod" can help you.
+
+
+How to use getmod.py
+--------------------
+
+Connect to KGTP before use the getmod.py.
+(gdb) source ~/kgtp/getmod.py
+Then this script will auto load the Linux kernel module's symbols to GDB.
+
+
+How to use getmod
+-----------------
+"getmod" is written by C so you can use it anywhere even if in an embedded
+environment.
+For example:
+
+--------------------------------------------------------------------------------
+#Following command save Linux Kernel module info to the file ~/tmp/mi in GDB
+#command format.
+sudo getmod >~/tmp/mi
+#in gdb part:
+(gdb) source ~/tmp/mi
+add symbol table from file
"/lib/modules/2.6.39-rc5+/kernel/fs/nls/nls_iso8859-1.ko" at
+ .text_addr = 0xf80de000
+ .note.gnu.build-id_addr = 0xf80de088
+ .exit.text_addr = 0xf80de074
+ .init.text_addr = 0xf8118000
+ .rodata.str1.1_addr = 0xf80de0ac
+ .rodata_addr = 0xf80de0c0
+ __mcount_loc_addr = 0xf80de9c0
+ .data_addr = 0xf80de9e0
+ .gnu.linkonce.this_module_addr = 0xf80dea00
+#After this GDB command, all the Linux Kernel module info is loaded into GDB.
+--------------------------------------------------------------------------------
+
+If you use remote debug or offline debug, maybe you need change the base
+directory. Following example is for it.
+
+--------------------------------------------------------------------------------
+#/lib/modules/2.6.39-rc5+/kernel is replaced to sudo ./getmod -r
/home/teawater/kernel/b26
+sudo ./getmod -r /home/teawater/kernel/b26 >~/tmp/mi
+--------------------------------------------------------------------------------
+
+
+
+Get GDB tracepoint commands
+---------------------------
+
+Please goto http://sourceware.org/gdb/current/onlinedocs/gdb/Tracepoints.html
+
+
+Get register info from Kernel
+-----------------------------
+
+The following is an example that records the value of all registers
+when "vfs_readdir" is called.
+
+--------------------------------------------------------------------------------
+(gdb) trace vfs_readdir
+Tracepoint 1 at 0xc01a1ac0: file
+/home/teawater/kernel/linux-2.6/fs/readdir.c, line 23.
+(gdb) actions
+Enter actions for tracepoint 1, one per line.
+End with a line saying just "end".
+>collect $reg
+>end
+(gdb) tstart
+(gdb) shell ls
+(gdb) tstop
+(gdb) tfind
+Found trace frame 0, tracepoint 1
+#0 0xc01a1ac1 in vfs_readdir (file=0xc5528d00, filler=0xc01a1900 <filldir64>,
+ buf=0xc0d09f90) at readdir.c:23
+23 readdir.c: No such file or directory.
+ in readdir.c
+(gdb) info reg
+eax 0xc5528d00 -984445696
+ecx 0xc0d09f90 -1060069488
+edx 0xc01a1900 -1072031488
+ebx 0xfffffff7 -9
+esp 0xc0d09f8c 0xc0d09f8c
+ebp 0x0 0x0
+esi 0x8061480 134616192
+edi 0xc5528d00 -984445696
+eip 0xc01a1ac1 0xc01a1ac1 <vfs_readdir+1>
+eflags 0x286 [ PF SF IF ]
+cs 0x60 96
+ss 0x8061480 134616192
+ds 0x7b 123
+es 0x7b 123
+fs 0x0 0
+gs 0x0 0
+(gdb) tfind
+Found trace frame 1, tracepoint 1
+0xc01a1ac1 23 in readdir.c
+(gdb) info reg
+eax 0xc5528d00 -984445696
+ecx 0xc0d09f90 -1060069488
+edx 0xc01a1900 -1072031488
+ebx 0xfffffff7 -9
+esp 0xc0d09f8c 0xc0d09f8c
+ebp 0x0 0x0
+esi 0x8061480 134616192
+edi 0xc5528d00 -984445696
+eip 0xc01a1ac1 0xc01a1ac1 <vfs_readdir+1>
+eflags 0x286 [ PF SF IF ]
+cs 0x60 96
+ss 0x8061480 134616192
+ds 0x7b 123
+es 0x7b 123
+fs 0x0 0
+gs 0x0 0
+--------------------------------------------------------------------------------
+
+
+
+Get the value of variable from Kernel
+-------------------------------------
+
+The following is an example that records the value of "jiffies_64" when the
+function "vfs_readdir" is called:
+
+--------------------------------------------------------------------------------
+(gdb) trace vfs_readdir
+Tracepoint 1 at 0xc01ed740: file
/home/teawater/kernel/linux-2.6/fs/readdir.c, line 24.
+(gdb) actions
+Enter actions for tracepoint 1, one per line.
+End with a line saying just "end".
+>collect jiffies_64
+>collect file->f_path.dentry->d_iname
+>end
+(gdb) tstart
+(gdb) shell ls
+arch drivers include kernel mm Module.symvers
security System.map virt
+block firmware init lib modules.builtin net
sound t vmlinux
+crypto fs ipc Makefile modules.order scripts
source usr vmlinux.o
+(gdb) tstop
+(gdb) tfind
+Found trace frame 0, tracepoint 1
+#0 0xc01ed741 in vfs_readdir (file=0xf4063000, filler=0xc01ed580
<filldir64>, buf=0xd6dfdf90)
+ at readdir.c:24
+24 {
+(gdb) p jiffies_64
+$1 = 4297248706
+(gdb) p file->f_path.dentry->d_iname
+$1 = "b26", '\000' <repeats 28 times>
+--------------------------------------------------------------------------------
+
+
+
+How to use use tracepoint condition
+-----------------------------------
+
+http://sourceware.org/gdb/current/onlinedocs/gdb/Tracepoint-Conditions.html
+Like breakpoints, we can set conditions on tracepoints. The speed of
+tracepoints is faster than breakpoints because KGTP can do all the condition
+checks.
+For example:
+
+------------------------------
+(gdb) trace handle_irq
+(gdb) condition 1 (irq == 47)
+------------------------------
+
+This action of tracepoint 1 will work only when irq number is 47.
+
+
+
+How to use trace state variables
+--------------------------------
+
+http://sourceware.org/gdb/current/onlinedocs/gdb/Trace-State-Variables.html
+Tracepoints have special variables. The variables can be traced directly,
+or used in tracepoint conditions.
+Note that just GDB 7.2.1 and later versions support use trace state variables
+directly, the old version of GDB can show the value of trace state variables
+through command "info tvariables".
+
+
+Simple trace state variables
+----------------------------
+
+Define a trace state variable $c.
+
+-------------------
+(gdb) tvariable $c
+-------------------
+
+Trace state variable $c is created with initial value 0.
+The following action uses $c to count how many irqs happened in the kernel.
+
+-----------------------------------------------------------------------
+(gdb) trace handle_irq
+(gdb) actions
+Enter actions for tracepoint 3, one per line.
+End with a line saying just "end".
+>collect $c #Save current value of $c to the trace frame buffer.
+>teval $c=$c+1 #Increase the $c.
+>end
+-----------------------------------------------------------------------
+
+Also, you can set a value of variable to trace state variable, but don't forget
+covert variable to "uint64_t".
+
+-----------------------------------------------------------------------
+>teval $c=(uint64_t)jiffies_64
+-----------------------------------------------------------------------
+
+You can get the current value of $c while the trace is running or stopped.
+
+----------------------------------
+(gdb) tstart
+(gdb) info tvariables
+$c 0 31554
+(gdb) p $c
+$5 = 33652
+(gdb) tstop
+(gdb) p $c
+$9 = 105559
+----------------------------------
+
+When using tfind, you can parse the trace frame buffer. If the value of a
+trace state variable is collected, you can parse it out.
+
+------------------------------
+(gdb) tstop
+(gdb) tfind
+(gdb) info tvariables
+$c 0 0
+(gdb) p $c
+$6 = 0
+(gdb) tfind 100
+(gdb) p $c
+$7 = 100
+------------------------------
+
+If need, the tracepoint action that access the simple trace state
variables will
+auto lock the spin lock for trace state variables. So it can handle race
+condition about trace state variables.
+The following example is OK even if it running a machine that have more than
+one CPU.
+-------------------------------------
+teval $c=$c+1
+-------------------------------------
+
+
+Per_cpu trace state variables
+-----------------------------
+
+Per_cpu trace state variables are special simple trace state variables.
+When tracepoint action access to it, it will access to this CPU special trace
+state variables.
+
+It have 2 advantages:
+1. The tracepoint actions that access to per_cpu trace state variables don't
+have the race conditon issue. So it don't need lock the spin lock for trace
+state variables. It is faster than simple trace state variables on multi-core
+machine.
+2. Write the action that count some CPU special thing with it is easier than
+simple trace state variables.
+
+To define per_cpu trace state variables, you need named it in
+format:
+"per_cpu_"+string+CPU_id
+or
+"pc_"+string+CPU_id
+Following example will define a series of per_cpu trace state variables
+in a 4 COREs CPU machine with string "count":
+-------------------------------------
+tvariable $pc_count0
+tvariable $pc_count1
+tvariable $pc_count2
+tvariable $pc_count3
+-------------------------------------
+
+You can use compatibility better way to do it:
+-----------------------------------------
+set $tmp=0
+while $tmp<$cpu_number
+ eval "tvariable $pc_count%d",$tmp
+ set $tmp=$tmp+1
+end
+-----------------------------------------
+
+Tracepoint action can access anyone of a series of per_cpu trace state
+variables. KGTP will auto access the one of CPU that it running on.
+For example:
+----------------------------------------
+trace vfs_read
+actions
+teval $pc_count0=$pc_count0+1
+end
+----------------------------------------
+These GDB commands define a tracepoint that count the times that call vfs_read
+of each CPU.
+
+
+Special trace state variables $current_task, $current_thread_info, $cpu_id,
+$dump_stack, $printk_level, $printk_format, $printk_tmp, $clock, $rdtsc,
+$hardirq_count, $softirq_count and $irq_count
+---------------------------------------------------------------------------
+
+KGTP special trace state variables $current_task, $current_thread_info,
+$cpu_id and $clock can very easy to access to some special value. You can see
+them when GDB connects to the KGTP. You can use them in tracepoint conditions
+or actions.
+Access $current_task in tracepoint condition and action will get that returns
+of get_current().
+Access $current_thread_info in tracepoint condition and action will get that
+returns of current_thread_info().
+Access $cpu_id in tracepoint condition and action will get that returns of
+smp_processor_id().
+Access $clock in tracepoint condition and action will get that returns of
+local_clock() that return the timestamp in nanoseconds.
+$rdtsc is only available on X86 and X86_64 architecture. Access it in anytime
+will get current value of TSC with instruction RDTSC.
+Access $hardirq_count in tracepoint condition and action will get that returns
+of hardirq_count().
+Access $softirq_count in tracepoint condition and action will get that returns
+of softirq_count().
+Access $irq_count in tracepoint condition and action will get that returns
+of irq_count().
+
+And KGTP has other special trace state variables $dump_stack, $printk_level,
+$printk_format and $printk_tmp. All of them output their values directly,
+as can be seen in "Howto let tracepoint output value directly".
+
+The following example counts in $c how many vfs_read calls that process 16663
+does and collects the struct thread_info of current task:
+
+--------------------------------------------------------------------------------
+(gdb) trace vfs_read if (((struct task_struct *)$current_task)->pid == 16663)
+(gdb) tvariable $c
+(gdb) actions
+Enter actions for tracepoint 4, one per line.
+End with a line saying just "end".
+>teval $c=$c+1
+>collect (*(struct thread_info *)$current_thread_info)
+>end
+(gdb) tstart
+(gdb) info tvariables
+Name Initial Current
+$c 0 184
+$current_task 0 <unknown>
+$current_thread_info 0 <unknown>
+$cpu_id 0 <unknown>
+(gdb) tstop
+(gdb) tfind
+(gdb) p *(struct thread_info *)$current_thread_info
+$10 = {task = 0xf0ac6580, exec_domain = 0xc07b1400, flags = 0, status
= 0, cpu = 1, preempt_count = 2, addr_limit = {
+ seg = 4294967295}, restart_block = {fn = 0xc0159fb0
<do_no_restart_syscall>, {{arg0 = 138300720, arg1 = 11,
+ arg2 = 1, arg3 = 78}, futex = {uaddr = 0x83e4d30, val = 11,
flags = 1, bitset = 78, time = 977063750,
+ uaddr2 = 0x0}, nanosleep = {index = 138300720, rmtp = 0xb,
expires = 335007449089}, poll = {
+ ufds = 0x83e4d30, nfds = 11, has_timeout = 1, tv_sec = 78,
tv_nsec = 977063750}}},
+ sysenter_return = 0xb77ce424, previous_esp = 0, supervisor_stack =
0xef340044 "", uaccess_err = 0}
+--------------------------------------------------------------------------------
+
+Another example shows how much sys_read() executes in each CPU.
+
+--------------------------------------
+tvariable $c0
+tvariable $c1
+trace sys_read
+ condition $bpnum ($cpu_id == 0)
+ commands
+ teval $c0=$c0+1
+ end
+trace sys_read
+ condition $bpnum ($cpu_id == 1)
+ commands
+ teval $c1=$c1+1
+ end
+info tvariables
+Name Initial Current
+$current_task 0 <unknown>
+$cpu_id 0 <unknown>
+$c0 0 3255
+$c1 0 1904
+--------------------------------------
+
+sys_read() execute 3255 times in cpu0 and 1904 times in cpu1.
+
+
+Special trace state variable $no_self_trace
+--------------------------------------------
+
+$no_self_trace is different with the special trace state variables in the
+previous section. It is used to control the behavior of tracepoint.
+If the action of a tracepoint include a command access to the $no_self_trace.
+The tracepoint will not trace anything if the current_task is the a KGTP self
+process (GDB, netcat, getframe or some others process that access to the
+interface of KGTP).
+For example, if we want trace vfs_read or something that have process context,
+and we don't want trace the operation of KGTP self process. Add following
+command to the action:
+
+--------------------------------------
+collect $no_self_trace
+--------------------------------------
+
+Please note that the code that doesn't have process context (Irq handler,
+softirq) doesn't need set this variable.
+
+
+Trace the function return with $kret
+------------------------------------
+
+Sometime, set the tracepoint to the end of function is hard because the Kernel
+is compiled with optimization. At this time, you can get help from $kret.
+
+$kret is a special trace state variable like $no_self_trace. When you set
+value of it inside the action of tracepoint, this tracepoint be set with
+kretprobe instead of kprobe. Then it can trace the end of this function.
+
+Following part is an example:
+
+------------------------------------------------------------------------------
+#"*(function_name)" format can make certain that GDB send the first address
+#of function to KGTP.
+trace *(vfs_read)
+action
+teval $kret=0
+#Following part you can set commands that you want.
+------------------------------------------------------------------------------
+
+
+Use $ignore_error and $last_errno to ignore the error of tstart
+---------------------------------------------------------------
+
+If KGTP got any error of tstart, this command will get fail.
+But sometime we need ignore this error and let KGTP keep work.
+For example: If you set tracepoint on the inline function spin_lock.
+This tracepoint will be set to a lot of addresses that some of them cannot
+be set kprobe. It will make tstart get fail. You can use "$ignore_error"
+ignore this error.
+And the last error number will available in "$last_errno".
+
+---------------------------------
+tvariable $ignore_error
+---------------------------------
+This command will open ignore.
+
+---------------------------------
+delete tvariable $ignore_error
+---------------------------------
+This command will close ignore.
+
+
+Use $cooked_clock and $cooked_rdtsc the time without KGTP used
+--------------------------------------------------------------
+
+Access these two trace state variables can get the time without KGTP used.
+Then we can get more close to really time that a part of code used even if the
+actions of tracepoint is very complex. They will be introduce in
+Cookbook (coming soon).
+
+
+Use $xtime_sec and $xtime_nsec get the timespec
+-----------------------------------------------
+
+Access these two trace state variables will return the time of day in
+a timespec that use getnstimeofday.
+$xtime_sec will access to the second part of a timespec.
+$xtime_nsec will access to the nanosecond part of a timespec.
+
+
+
+How to use performance counters
+-------------------------------
+
+Performance counters are special hardware registers available on most modern
+CPUs. These registers count the number of certain types of hw events: such as
+instructions executed, cachemisses suffered, or branches
mis-predicted - without
+slowing down the kernel or applications. These registers can also trigger
+interrupts when a threshold number of events have passed - and can thus be
+used to profile the code that runs on that CPU.
+
+The Linux Performance Counter subsystem called perf event can get the value of
+performance counter. You can access it through KGTP perf event trace state
+variables.
+
+Please goto read the file tools/perf/design.txt in Linux Kernel to get more
+info about perf event.
+
+
+Define a perf event trace state variable
+----------------------------------------
+
+Access an performance counter need define following trace state variable:
+-----------------------------------------------------------------------------
+"pe_cpu_"+tv_name Define the the CPU id of the performance counter.
+"pe_type_"+tv_name Define the the type of the performance counter.
+"pe_config_"+tv_name Define the the config of the performance counter.
+"pe_en_"+tv_name This the switch to enable or disable the performance
+ counter.
+ The performance counter is disable in default.
+"pe_val_"+tv_name Access this variable can get the value of the
+ performance counter.
+-----------------------------------------------------------------------------
+
+
+Define a per_cpu perf event trace state variable
+------------------------------------------------
+
+Define a per_cpu perf event trace state variable is same with define
+"Per_cpu_trace_state_variables".
+
+------------------------------------------
+"pc_pe_"+perf_event type+string+CPU_id
+------------------------------------------
+
+Note that if you define a per_cpu perf event trace state variable, you will not
+need define the "pe_cpu_" because KGTP already get it from per_cpu id.
+
+
+The perf event type and config
+------------------------------
+
+The type of perf event can be:
+-----------------------------------------------------------
+0 PERF_TYPE_HARDWARE
+1 PERF_TYPE_SOFTWARE
+2 PERF_TYPE_TRACEPOINT
+3 PERF_TYPE_HW_CACHE
+4 PERF_TYPE_RAW
+5 PERF_TYPE_BREAKPOINT
+-----------------------------------------------------------
+
+If the type is 0(PERF_TYPE_HARDWARE), the config can be:
+-----------------------------------------------------------
+0 PERF_COUNT_HW_CPU_CYCLES
+1 PERF_COUNT_HW_INSTRUCTIONS
+2 PERF_COUNT_HW_CACHE_REFERENCES
+3 PERF_COUNT_HW_CACHE_MISSES
+4 PERF_COUNT_HW_BRANCH_INSTRUCTIONS
+5 PERF_COUNT_HW_BRANCH_MISSES
+6 PERF_COUNT_HW_BUS_CYCLES
+7 PERF_COUNT_HW_STALLED_CYCLES_FRONTEND
+8 PERF_COUNT_HW_STALLED_CYCLES_BACKEND
+-----------------------------------------------------------
+
+If the type is 3(PERF_TYPE_HW_CACHE), the config need to divide to 3 parts:
+First one is cache id, it need be << 0 before set to config:
+-----------------------------------------------------------
+0 PERF_COUNT_HW_CACHE_L1D
+1 PERF_COUNT_HW_CACHE_L1I
+2 PERF_COUNT_HW_CACHE_LL
+3 PERF_COUNT_HW_CACHE_DTLB
+4 PERF_COUNT_HW_CACHE_ITLB
+5 PERF_COUNT_HW_CACHE_BPU
+-----------------------------------------------------------
+Second one is cache op id, it need be << 8 before set to config:
+-----------------------------------------------------------
+0 PERF_COUNT_HW_CACHE_OP_READ
+1 PERF_COUNT_HW_CACHE_OP_WRITE
+2 PERF_COUNT_HW_CACHE_OP_PREFETCH
+-----------------------------------------------------------
+Last one is cache op result id, it need be << 16 before set to config:
+-----------------------------------------------------------
+0 PERF_COUNT_HW_CACHE_RESULT_ACCESS
+1 PERF_COUNT_HW_CACHE_RESULT_MISS
+-----------------------------------------------------------
+If you want get the perf count of PERF_COUNT_HW_CACHE_L1I(1),
+PERF_COUNT_HW_CACHE_OP_WRITE(1) and PERF_COUNT_HW_CACHE_RESULT_MISS(1),
+you can use:
+-----------------------------------------------------------
+tvariable $pe_config_cache=1 | (1 << 8) | (1 << 16)
+-----------------------------------------------------------
+
+tools/perf/design.txt in Linux Kernel have more info about type and config
+of perf event.
+
+
+Enable and disable all the perf event in a CPU with $pc_pe_en
+-------------------------------------------------------------
+
+I think the best way that count a part of code with performance counters is
+enable all the count in the begin of the code and disable all of them in the
+end. You can do it with "pe_en_". But if you have a lot of perf event trace
+state variables. That will make the tracepoint action very big. $pc_pe_en is
+for this issue.
+
+You can enable all the perf event trace state variables in current CPU with
+following action:
+--------------------
+teval $pc_pe_en=1
+--------------------
+Disable them with set $pc_pe_en to 0.
+--------------------
+teval $pc_pe_en=0
+--------------------
+
+
+GDB scripts to help with set and get the perf event trace state variables
+-------------------------------------------------------------------------
+
+Following is a GDB script define two commands dpe and spe to help define and
+show the perf event trace state variables.
+You can put it to the ~/.gdbinit or your tracepoint script. Then you can use
+this two commands in GDB directly.
+-------------------------------------------------------------------------------
+define dpe
+ if ($argc < 2)
+ printf "Usage: dpe pe_type pe_config [enable]\n"
+ end
+ if ($argc >= 2)
+ set $tmp=0
+ while $tmp<$cpu_number
+ eval "tvariable $pc_pe_type_%d%d_%d=%d",$arg0, $arg1, $tmp, $arg0
+ eval "tvariable $pc_pe_config_%d%d_%d=%d",$arg0, $arg1, $tmp, $arg1
+ eval "tvariable $pc_pe_val_%d%d_%d=0",$arg0, $arg1, $tmp
+ if ($argc >= 3)
+ eval "tvariable $pc_pe_en_%d%d_%d=%d",$arg0, $arg1, $tmp, $arg2
+ end
+ set $tmp=$tmp+1
+ end
+ end
+end
+
+document dpe
+Usage: dpe pe_type pe_config [cpu_id]
+end
+
+define spe
+ if ($argc != 2 && $argc != 3)
+ printf "Usage: spe pe_type pe_config [cpu_id]\n"
+ end
+ if ($argc == 2)
+ set $tmp=0
+ while $tmp<$cpu_number
+ eval "printf \"$pc_pe_val_%%d%%d_%%d=%%ld\\n\",$arg0, $arg1,
$tmp, $pc_pe_val_%d%d_%d", $arg0, $arg1, $tmp
+ set $tmp=$tmp+1
+ end
+ end
+ if ($argc == 3)
+ eval "printf \"$pc_pe_val_%%d%%d_%%d=%%ld\\n\",$arg0, $arg1,
$tmp, $pc_pe_val_%d%d_%d", $arg0, $arg1, $arg2
+ end
+end
+
+document spe
+Usage: spe pe_type pe_config [cpu_id]
+end
+-------------------------------------------------------------------------------
+
+Following is an example to use it get the performance counters of
function tcp_v4_rcv:
+-------------------------------------------------------------------------------
+#Connect to kgtp
+target remote /sys/kernel/debug/gtp
+#Define 3 pe tvs for PERF_COUNT_HW_CPU_CYCLES,
PERF_COUNT_HW_CACHE_MISSES and PERF_COUNT_HW_BRANCH_MISSES.
+dpe 0 0
+dpe 0 3
+dpe 0 5
+#enable the performance counters of this CPU in the begin of this function.
+trace tcp_v4_rcv
+ action
+ teval $pc_pe_en=1
+ end
+#$kret make this hanler the end of function tcp_v4_rcv.
+trace *(tcp_v4_rcv)
+ action
+ teval $kret=0
+ #disable all performance counters of this CPU
+ teval $pc_pe_en=0
+ #Access the per cpu perf event tv will access to the current cpu pe tv.
+ collect $pc_pe_val_00_0
+ collect $pc_pe_val_03_0
+ collect $pc_pe_val_05_0
+ #Set all the pe tv to 0
+ teval $pc_pe_val_00_0=0
+ teval $pc_pe_val_03_0=0
+ teval $pc_pe_val_05_0=0
+ end
+tstart
+#Wait some time that current pc receive some tcp package.
+tstop
+tfind
+spe 0 0 $cpu_id
+$pc_pe_val_00_2=12676
+spe 0 3 $cpu_id
+$pc_pe_val_03_2=7
+spe 0 5 $cpu_id
+$pc_pe_val_05_2=97
+-------------------------------------------------------------------------------
+
+
+
+Show all the traced data of current frame
+-----------------------------------------
+
+--------------------------------------------------------------------------------
+(gdb) tdump
+Data collected at tracepoint 1, trace frame 0:
+$cr = void
+file->f_path.dentry->d_iname =
"gtp\000.google.chrome.g05ZYO\000\235", <incomplete sequence \364>
+jiffies_64 = 4319751455
+--------------------------------------------------------------------------------
+
+
+
+Get backtrace info(stack dump) from Kernel
+------------------------------------------
+
+We can get a backtrace(stack dump) by collecting the stack.
+In x86_32, following action command will collect 512 bytes of stack.
+
+-----------------------------------
+collect *(unsigned char *)$esp@512
+-----------------------------------
+
+In x86_64, following command will collect 512 bytes of stack.
+
+-----------------------------------
+collect *(unsigned char *)$rsp@512
+-----------------------------------
+
+In MIPS or ARM, following command will collect 512 bytes of stack.
+----------------------------------
+collect *(unsigned char *)$sp@512
+-----------------------------------
+
+Following part is an example about howto backtrace in x86_64:
+
+--------------------------------------------------------------------------------
+(gdb) trace vfs_readdir
+Tracepoint 1 at 0xffffffff8113f7fc: file
/home/teawater/kernel/linux-2.6/fs/readdir.c, line 24.
+(gdb) actions
+Enter actions for tracepoint 1, one per line.
+End with a line saying just "end".
+>collect *(unsigned char *)$rsp@512
+>end
+(gdb) tstart
+(gdb) shell ls
+2 block firmware i ipc Makefile
modules.order scripts source t~ vmlinux
+a.out crypto fs include kernel mm
Module.symvers security System.map usr vmlinux.o
+arch drivers gdb.txt init lib modules.builtin net
sound t virt
+(gdb) tstop
+(gdb) tfind
+Found trace frame 0, tracepoint 1
+#0 0xffffffff8113f7fd in vfs_readdir (file=0xffff880075f00780,
filler=0xffffffff8113f630 <filldir>, buf=0xffff880005785f38)
+ at ./linux-2.6/fs/readdir.c:24
+24 {
+(gdb) bt
+#0 0xffffffff8113f7fd in vfs_readdir (file=0xffff880075f00780,
filler=0xffffffff8113f630 <filldir>, buf=0xffff880005785f38)
+ at ./linux-2.6/fs/readdir.c:24
+#1 0xffffffff8113fa14 in sys_getdents (fd=<value optimized out>,
dirent=0x801108, count=32768)
+ at ./linux-2.6/fs/readdir.c:214
+#2 0xffffffff8100af42 in ?? () at ./linux-2.6/arch/x86/kernel/entry_64.S:487
+--------------------------------------------------------------------------------
+
+
+
+Howto let tracepoint output value directly
+------------------------------------------
+
+In the previous parts, you may understand that to get a value from
Linux Kernel,
+you need to use a tracepoint "collect" action to save the value to the
+tracepoint frame and use the GDB command "tfind" to parse the value from the
+frame data.
+But we want get the value directly sometimes, so KGTP supports two ways to
+output values directly.
+
+
+Output stack dump directly
+--------------------------
+KGTP has special trace state variable $dump_stack, "collect" it will let Linux
+Kernel output stack dump directly.
+Following example lets Linux Kernel show the stack dump of vfs_readdir:
+
+--------------------------------------------------------------------------------
+trace vfs_readdir
+ commands
+ collect $dump_stack
+ end
+--------------------------------------------------------------------------------
+
+Then your kernel will printk like:
+
+--------------------------------------------------------------------------------
+[22779.208064] gtp 1:Pid: 441, comm: python Not tainted 2.6.39-rc3+ #46
+[22779.208068] Call Trace:
+[22779.208072] [<fe653cca>] gtp_get_var+0x4a/0xa0 [gtp]
+[22779.208076] [<fe653d79>] gtp_collect_var+0x59/0xa0 [gtp]
+[22779.208080] [<fe655974>] gtp_action_x+0x1bb4/0x1dc0 [gtp]
+[22779.208084] [<c05b6408>] ? _raw_spin_unlock+0x18/0x40
+[22779.208088] [<c023f152>] ? __find_get_block_slow+0xd2/0x160
+[22779.208091] [<c01a8c56>] ? delayacct_end+0x96/0xb0
+[22779.208100] [<c023f404>] ? __find_get_block+0x84/0x1d0
+[22779.208103] [<c05b6408>] ? _raw_spin_unlock+0x18/0x40
+[22779.208106] [<c02e0838>] ? find_revoke_record+0xa8/0xc0
+[22779.208109] [<c02e0c45>] ? jbd2_journal_cancel_revoke+0xd5/0xe0
+[22779.208112] [<c02db51f>] ? __jbd2_journal_temp_unlink_buffer+0x2f/0x110
+[22779.208115] [<fe655c4c>] gtp_kp_pre_handler+0xcc/0x1c0 [gtp]
+[22779.208118] [<c05b8a88>] kprobe_exceptions_notify+0x3d8/0x440
+[22779.208121] [<c05b7d54>] ? hw_breakpoint_exceptions_notify+0x14/0x180
+[22779.208124] [<c05b95eb>] ? sub_preempt_count+0x7b/0xb0
+[22779.208126] [<c0227ac5>] ? vfs_readdir+0x15/0xb0
+[22779.208128] [<c0227ac4>] ? vfs_readdir+0x14/0xb0
+[22779.208131] [<c05b9743>] notifier_call_chain+0x43/0x60
+[22779.208134] [<c05b9798>] __atomic_notifier_call_chain+0x38/0x50
+[22779.208137] [<c05b97cf>] atomic_notifier_call_chain+0x1f/0x30
+[22779.208140] [<c05b980d>] notify_die+0x2d/0x30
+[22779.208142] [<c05b71c5>] do_int3+0x35/0xa0
+--------------------------------------------------------------------------------
+
+
+Switch collect to output the value directly
+-------------------------------------------
+
+KGTP has special trace state variables $printk_level, $printk_format
+and $printk_tmp to support this function.
+
+$printk_level: if its value is 8 (this is the default value), "collect" action
+will save value to the tracepoint frame in the simple behavior.
+If its value is 0-7, "collect" will output the value through "printk" directly,
+and value will be the level of printk. The level is:
+0 KERN_EMERG system is unusable
+1 KERN_ALERT action must be taken immediately
+2 KERN_CRIT critical conditions
+3 KERN_ERR error conditions
+4 KERN_WARNING warning conditions
+5 KERN_NOTICE normal but significant condition
+6 KERN_INFO informational
+7 KERN_DEBUG debug-level messages
+
+$printk_format, collect printk will output value in the format that is set
+by it.
+The format is:
+0 This is the default value.
+ If the size of collect value is 1, 2, 4 or 8, it will be output as an
+ unsigned decimal.
+ If not, it will be output as a hexadecimal string.
+1 Output value in signed decimal.
+2 Output value in unsigned decimal.
+3 Output value in unsigned hexadecimal.
+4 Output value as a string.
+5 Output value as a hexadecimal string.
+
+$printk_tmp, to output the value of global variable need set to it first.
+
+Following example shows a count number, pid, jiffies_64 and the file name
+that call vfs_readdir:
+
+--------------------------------------------------------------------------------
+tvariable $c
+trace vfs_readdir
+ commands
+ teval $printk_level=0
+ collect $c=$c+1
+ collect ((struct task_struct *)$current_task)->pid
+ collect $printk_tmp=jiffies_64
+ teval $printk_format=4
+ collect file->f_path.dentry->d_iname
+ end
+--------------------------------------------------------------------------------
+
+Then your kernel will printk like:
+
+--------------------------------------------------------------------------------
+gtp 1:$c=$c+1=41
+gtp 1:((struct task_struct *)$current_task)->pid=12085
+gtp 1:$printk_tmp=jiffies_64=4322021438
+gtp 1:file->f_path.dentry->d_iname=b26
+gtp 1:$c=$c+1=42
+gtp 1:((struct task_struct *)$current_task)->pid=12085
+gtp 1:$printk_tmp=jiffies_64=4322021438
+gtp 1:file->f_path.dentry->d_iname=b26
+--------------------------------------------------------------------------------
+
+"gtp 1" means that it was output by tracepoint 1.
+
+
+Use printf command in actions
+-----------------------------
+
+This way have a trouble is GDB is still not accept the patch that make
+tracepoint support printf, So if you want use it, you need patch the patch in
+http://sourceware.org/ml/gdb-patches/2011-03/msg00022.html and build your
+GDB yourself.
+
+Following example show the a count number, pid and the file name that call
+vfs_readdir:
+
+--------------------------------------------------------------------------------
+tvariable $c
+trace vfs_readdir
+ commands
+ printf "<0>%d pid=%d name:%s\n", $c=$c+1, ((struct task_struct
*)$current_task)->pid, file->f_path.dentry->d_iname
+ end
+--------------------------------------------------------------------------------
+
+Then your kernel will printk like:
+
+--------------------------------------------------------------------------------
+1 pid=10888 name:bin
+2 pid=10888 name:bin
+3 pid=10893 name:teawater
+4 pid=10893 name:teawater
+--------------------------------------------------------------------------------
+
+Like what we use with printk in Linux Kernel, please add kernel loglevel at the
+beginning and add "\n" in the end.
+The kernel loglevel is:
+KERN_EMERG "<0>" system is unusable
+KERN_ALERT "<1>" action must be taken immediately
+KERN_CRIT "<2>" critical conditions
+KERN_ERR "<3>" error conditions
+KERN_WARNING "<4>" warning conditions
+KERN_NOTICE "<5>" normal but significant condition
+KERN_INFO "<6>" informational
+KERN_DEBUG "<7>" debug-level messages
+
+
+
+Get status of KGTP from Kernel
+------------------------------
+Please use GDB command "tstatus"
+
+
+
+Set the trace buffer into a circular buffer
+-------------------------------------------
+http://sourceware.org/gdb/current/onlinedocs/gdb/Starting-and-Stopping-Trace-Experiments.html
+The frame buffer is not a circular buffer by default. When the buffer is full,
+the tracepoint will stop.
+
+-----------------------------
+set circular-trace-buffer on
+-----------------------------
+
+Set frame buffer to a circular buffer. When the buffer is full, it will auto
+discard traceframes (oldest first) and keep trace.
+
+
+
+Do not stop tracepoint when the GDB disconnects
+---------------------------------------------------
+
+http://sourceware.org/gdb/current/onlinedocs/gdb/Starting-and-Stopping-Trace-Experiments.html
+KGTP will stop and delete the trace frame when GDB disconnects with it by
+default.
+
+----------------------------
+set disconnected-tracing on
+----------------------------
+will open the KGTP disconnect-trace. After that, when GDB disconnects with
+KGTP, KGTP will not stop trace. And after GDB reconnects to KGTP, it can keep
+control of KGTP like nothing happened.
+
+
+
+Howto show a variable whose value has been optimized away
+---------------------------------------------------------
+
+Sometimes, GDB will output some value like:
+
+-------------------------------------------
+inode has been optimized out of existence.
+res has been optimized out of existence.
+-------------------------------------------
+
+That is because value of inode and res is optimized. Linux Kernel is built
+with -O2 so you will get this trouble sometimes.
+There are 2 ways to handle it:
+
+Linux kernel "Compile with almost no optimization" patch
+--------------------------------------------------------
+If you do not care about the speed when you debug the Kernel, you can use the
+patch co.patch include in the source of KGTP or goto
+http://code.google.com/p/kgtp/downloads/detail?name=co.patch to get it. It add
+a option in "Kernel hacking" called "Compile with almost no optimization". It
+will make kernel be built without -O2. It support x86_32, x86_64 and arm.
+
+Update your GCC
+---------------
+The VTA branch (http://gcc.gnu.org/wiki/Var_Tracking_Assignments) was merged
+for GCC 4.5. This helps a lot with generating dwarf for previously
+"optimized out" values.
+
+
+
+Offline debug
+-------------
+
+In the PC that can run the GDB:
+Change the "target remote XXXX" to
+
+------------------------------------------
+(gdb) target remote | perl ./getgtprsp.pl
+------------------------------------------.
+
+After that, set tracepoint and start it as usual:
+
+--------------------------------------------------------------------------------
+(gdb) trace vfs_readdir
+Tracepoint 1 at 0xffffffff8114f3c0: file
/home/teawater/kernel/linux-2.6/fs/readdir.c, line 24.
+(gdb) actions
+Enter actions for tracepoint 1, one per line.
+End with a line saying just "end".
+#If your GDB support tracepoint "printf" (see "Howto use tracepoint
printf"), use it to show the value directly is better.
+>collect $reg
+>end
+(gdb) tstart
+(gdb) stop
+(gdb) quit
+--------------------------------------------------------------------------------
+
+Then you can find files gtpstart and gtpstop. Copy it to the machine that you
+want to debug.
+
+
+In the debugged machine after insmod the gtp.ko:
+Start the tracepoint:
+
+------------------------------------
+cat gtpstart > /sys/kernel/debug/gtp
+------------------------------------
+
+Stop the tracepoint:
+
+-----------------------------------
+cat gtpstop > /sys/kernel/debug/gtp
+-----------------------------------
+
+You can let Linux Kernel show the value directly, please see "Howto let
+tracepoint output value directly".
+
+If you want to save the value to the trace frame and parse later, you can use
+file "/sys/kernel/debug/gtpframe" that has the trace frame. Copy it to the PC
+that has GDB.
+Please note that some "cp" cannot handle it very well, please use
+"cat /sys/kernel/debug/gtpframe > ./gtpframe" to copy it.
+
+In the PC that can run the GDB:
+
+--------------------------------------------------------------------------------
+(gdb) target tfile ./gtpframe
+Tracepoint 1 at 0xffffffff8114f3dc: file
/home/teawater/kernel/linux-2.6/fs/readdir.c, line 24.
+Created tracepoint 1 for target's tracepoint 1 at 0xffffffff8114f3c0.
+(gdb) tfind
+Found trace frame 0, tracepoint 1
+#0 vfs_readdir (file=0xffff880036e8f300, filler=0xffffffff8114f240
<filldir>, buf=0xffff880001e5bf38)
+ at readdir.c:24
+24 {
+--------------------------------------------------------------------------------
--- /dev/null
+++ b/Documentation/trace/gtp_quickstart.txt
@@ -0,0 +1,250 @@
+ Linux Kernel GDB tracepoint module (KGTP) quick start
+ =====================================================
+ By Hui Zhu <teawater@...il.com>
+ https://code.google.com/p/kgtp/wiki/Quickstart
+ 2011-09-12
+
+Table of contents
+-----------------
+Ubuntu
+Fedora
+
+
+
+
+Ubuntu
+------
+
+Install GDB for KGTP
+--------------------
+
+This GDB's filename is different with the current GDB that you are using.
+So please don't worry that it affect current GDB that your are using.
+
+For the Ubuntu 10.04 or later, running the following line at a terminal:
+sudo add-apt-repository ppa:teawater/gdb-$(lsb_release -rs)
+sudo apt-get update
+sudo apt-get install gdb-release
+
+For the Ubuntu older than 10.04, please go to https://code.google.com/p/gdbt/
+get howto install GDB for KGTP from source.
+
+
+
+Install Linux kernel packages that KGTP need
+--------------------------------------------
+
+Please ignore this section if the Linux kernel of your system is built by
+yourself.
+
+Install the Linux kernel debug image
+------------------------------------
+
+Add debug source to the sources list of Ubuntu
+----------------------------------------------
+
+Create an /etc/apt/sources.list.d/ddebs.list by running the following line at
+a terminal:
+echo "deb http://ddebs.ubuntu.com $(lsb_release -cs) main restricted
universe multiverse" | \
+sudo tee -a /etc/apt/sources.list.d/ddebs.list
+
+Stable releases (not alphas and betas) require three more lines adding to the
+same file, which is done by the following terminal command:
+echo "deb http://ddebs.ubuntu.com $(lsb_release -cs)-updates main
restricted universe multiverse
+deb http://ddebs.ubuntu.com $(lsb_release -cs)-security main
restricted universe multiverse
+deb http://ddebs.ubuntu.com $(lsb_release -cs)-proposed main
restricted universe multiverse" | \
+sudo tee -a /etc/apt/sources.list.d/ddebs.list
+
+Import the debug symbol archive signing key:
+sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys 428D7C01
+
+Then run:
+sudo apt-get update
+
+Get Linux kernel debug image
+----------------------------
+sudo apt-get install linux-image-$(uname -r)-dbgsym
+
+
+Install the Linux kernel headers
+--------------------------------
+
+Please ignore this section if the Linux kernel of your system is built by
+yourself.
+sudo apt-get install linux-headers-generic
+
+
+Install the Linux kernel source
+-------------------------------
+
+Install the source package:
+sudo apt-get install linux-source
+
+Uncompress the source package:
+sudo mkdir -p /build/buildd/
+sudo tar vxjf /usr/src/linux-source-$(uname -r | sed
's/-.*//').tar.bz2 -C /build/buildd/
+sudo mv /build/buildd/linux-source-$(uname -r | sed 's/-.*//')
/build/buildd/linux-$(uname -r | sed 's/-.*//')
+
+
+
+Install GCC
+-----------
+
+sudo apt-get install gcc
+
+
+
+Get and build KGTP
+------------------
+
+Install subversion:
+sudo apt-get install subversion
+
+Get the source of KGTP with subversion and put it to directory "kgtp":
+svn checkout https://kgtp.googlecode.com/svn/trunk kgtp
+
+Build KGTP:
+cd kgtp
+make
+
+
+
+Use KGTP
+--------
+
+Mount the sysfs and debugfs:
+sudo mount -t sysfs none /sys/
+sudo mount -t debugfs none /sys/kernel/debug/
+
+Insert the KGTP module to the current Linux Kernel:
+cd kgtp
+sudo insmod gtp.ko
+
+Use GDB connect to KGTP:
+sudo gdb-release /usr/lib/debug/boot/vmlinux-$(uname -r)
+(gdb) target remote /sys/kernel/debug/gtp
+
+Do a very simple trace:
+(gdb) trace vfs_readdir
+Tracepoint 1 at 0xc02289f0: file
/build/buildd/linux-2.6.35/fs/readdir.c, line 23.
+(gdb) actions
+Enter actions for tracepoint 1, one per line.
+End with a line saying just "end".
+>collect $reg
+>end
+(gdb) tstart
+(gdb) shell ls
+vmlinux-2.6.35-30-generic
+(gdb) tstop
+(gdb) tfind
+Found trace frame 0, tracepoint 1
+#0 vfs_readdir (file=0x0, filler=0x163d8ae3, buf=0x18c0) at
/build/buildd/linux-2.6.35/fs/readdir.c:23
+23 {
+
+
+
+End
+---
+
+Now, you can begin to rock and roll your Linux kernel with KGTP and GDB.
+Please go to see gtp.txt to get more message about howto use KGTP.
+
+
+
+
+Fedora
+------
+
+Install GDB for KGTP
+--------------------
+
+Please go to https://code.google.com/p/gdbt/ get howto install GDB for KGTP
+from source.
+
+
+
+Install Linux kernel packages that KGTP need
+--------------------------------------------
+
+Please ignore this section if the Linux kernel of your system is built
+by yourself.
+
+
+Install the Linux kernel debug image
+------------------------------------
+
+sudo yum --enablerepo=fedora-debuginfo install kernel-debuginfo
+
+
+Install the Linux kernel devel package
+--------------------------------------
+
+sudo yum install kernel-devel-$(uname -r)
+
+
+
+Install GCC
+-----------
+
+sudo yum install gcc
+
+
+
+Get and build KGTP
+------------------
+
+Install subversion:
+sudo yum install subversion
+
+Get the source of KGTP with subversion and put it to directory "kgtp":
+svn checkout https://kgtp.googlecode.com/svn/trunk kgtp
+
+Build KGTP:
+cd kgtp
+make
+
+
+
+Use KGTP
+--------
+
+Mount the sysfs and debug fs:
+sudo mount -t sysfs none /sys/
+sudo mount -t debugfs none /sys/kernel/debug/
+
+Insert the KGTP module to the current Linux Kernel:
+cd kgtp
+sudo insmod gtp.ko
+
+Use GDB connect to KGTP:
+sudo gdb-release /usr/lib/debug/lib/modules/$(uname -r)/vmlinux
+(gdb) target remote /sys/kernel/debug/gtp
+
+Do a very simple trace:
+(gdb) trace vfs_readdir
+Tracepoint 1 at 0xffffffff8110ec9b: file fs/readdir.c, line 23.
+(gdb) actions
+Enter actions for tracepoint 1, one per line.
+End with a line saying just "end".
+>collect $reg
+>end
+(gdb) tstart
+(gdb) shell ls
+co.patch getframe getmod.c gtp.mod.c gtp.txt
perf_event.c
+dkms.conf getframe.c getmod.py gtp.mod.o
Makefile ring_buffer.c
+dkms_others_install.sh getgtprsp.pl gtp.c gtp.o
modules.order ring_buffer.h
+dkms_others_uninstall.sh getmod gtp.ko gtp.patch Module.symvers
+(gdb) tstop
+(gdb) tfind
+Found trace frame 0, tracepoint 1
+#0 vfs_readdir (file=0xffff880019d3df00, filler=0xffffffff8110eb16
<filldir>, buf=0xffff880003b39f38)
+ at fs/readdir.c:23
+23 {
+
+
+
+End
+---
+
+Now, you can begin to rock and roll your Linux kernel with KGTP and GDB.
+Please go to HOWTO to get more message about howto use KGTP.
--- /dev/null
+++ b/arch/arm/include/asm/gtp.h
@@ -0,0 +1,13 @@
+#ifndef _ASM_ARM_GTP_H_
+#define _ASM_ARM_GTP_H_
+
+#define ULONGEST uint64_t
+#define LONGEST int64_t
+#define CORE_ADDR unsigned long
+
+#define GTP_REGS_PC(regs) ((regs)->uregs[15])
+
+#define GTP_REG_ASCII_SIZE 336
+#define GTP_REG_BIN_SIZE 168
+
+#endif
--- /dev/null
+++ b/arch/mips/include/asm/gtp.h
@@ -0,0 +1,18 @@
+#ifndef _ASM_MIPS_GTP_H_
+#define _ASM_MIPS_GTP_H_
+
+#define ULONGEST uint64_t
+#define LONGEST int64_t
+#define CORE_ADDR unsigned long
+
+#define GTP_REGS_PC(regs) ((regs)->cp0_epc)
+
+#ifdef CONFIG_32BIT
+#define GTP_REG_ASCII_SIZE 304
+#define GTP_REG_BIN_SIZE 152
+#else
+#define GTP_REG_ASCII_SIZE 608
+#define GTP_REG_BIN_SIZE 304
+#endif
+
+#endif
--- /dev/null
+++ b/arch/x86/include/asm/gtp.h
@@ -0,0 +1,18 @@
+#ifndef _ASM_X86_GTP_H_
+#define _ASM_X86_GTP_H_
+
+#define ULONGEST uint64_t
+#define LONGEST int64_t
+#define CORE_ADDR unsigned long
+
+#define GTP_REGS_PC(regs) ((regs)->ip)
+
+#ifdef CONFIG_X86_32
+#define GTP_REG_ASCII_SIZE 128
+#define GTP_REG_BIN_SIZE 64
+#else
+#define GTP_REG_ASCII_SIZE 296
+#define GTP_REG_BIN_SIZE 148
+#endif
+
+#endif
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -1134,9 +1134,12 @@ extern void perf_output_copy(struct perf
const void *buf, unsigned int len);
extern int perf_swevent_get_recursion_context(void);
extern void perf_swevent_put_recursion_context(int rctx);
+extern int __perf_event_enable(void *info);
+extern int __perf_event_disable(void *info);
extern void perf_event_enable(struct perf_event *event);
extern void perf_event_disable(struct perf_event *event);
extern void perf_event_task_tick(void);
+extern void perf_event_set(struct perf_event *event, u64 val);
#else
static inline void
perf_event_task_sched_in(struct task_struct *task) { }
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -1187,7 +1187,7 @@ retry:
/*
* Cross CPU call to disable a performance event
*/
-static int __perf_event_disable(void *info)
+int __perf_event_disable(void *info)
{
struct perf_event *event = info;
struct perf_event_context *ctx = event->ctx;
@@ -1224,6 +1224,7 @@ static int __perf_event_disable(void *in
return 0;
}
+EXPORT_SYMBOL_GPL(__perf_event_disable);
/*
* Disable a event.
@@ -1279,6 +1280,7 @@ retry:
}
raw_spin_unlock_irq(&ctx->lock);
}
+EXPORT_SYMBOL_GPL(perf_event_disable);
static void perf_set_shadow_time(struct perf_event *event,
struct perf_event_context *ctx,
@@ -1633,7 +1635,7 @@ static void __perf_event_mark_enabled(st
/*
* Cross CPU call to enable a performance event
*/
-static int __perf_event_enable(void *info)
+int __perf_event_enable(void *info)
{
struct perf_event *event = info;
struct perf_event_context *ctx = event->ctx;
@@ -1697,6 +1699,7 @@ unlock:
return 0;
}
+EXPORT_SYMBOL_GPL(__perf_event_enable);
/*
* Enable a event.
@@ -1763,6 +1766,7 @@ retry:
out:
raw_spin_unlock_irq(&ctx->lock);
}
+EXPORT_SYMBOL_GPL(perf_event_enable);
int perf_event_refresh(struct perf_event *event, int refresh)
{
@@ -3164,6 +3168,14 @@ static void perf_event_reset(struct perf
perf_event_update_userpage(event);
}
+void perf_event_set(struct perf_event *event, u64 val)
+{
+ (void)perf_event_read(event);
+ local64_set(&event->count, val);
+ perf_event_update_userpage(event);
+}
+EXPORT_SYMBOL_GPL(perf_event_set);
+
/*
* Holding the top-level event's child_mutex means that any
* descendant process that has inherited this event will block
--- a/lib/Kconfig.debug
+++ b/lib/Kconfig.debug
@@ -1256,6 +1256,17 @@ config ASYNC_RAID6_TEST
If unsure, say N.
+config GTP
+ tristate "GDB tracepoint support"
+ depends on X86 || ARM || MIPS
+ depends on KPROBES
+ depends on DEBUG_FS
+ select RING_BUFFER
+ ---help---
+ Supply GDB tracepoint interface in /sys/kernel/debug/gtp.
+ See Documentation/trace/gtp.txt or
+ https://code.google.com/p/kgtp/wiki/HOWTO for more info.
+
source "samples/Kconfig"
source "lib/Kconfig.kgdb"
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -117,6 +117,8 @@ obj-$(CONFIG_CORDIC) += cordic.o
obj-$(CONFIG_LLIST) += llist.o
+obj-$(CONFIG_GTP) += gtp.o
+
hostprogs-y := gen_crc32table
clean-files := crc32table.h
--- /dev/null
+++ b/lib/gtp.c
@@ -0,0 +1,7323 @@
+/*
+ * Kernel GDB tracepoint module.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright(C) Hui Zhu (teawater@...il.com), 2010, 2011
+ */
+
+#include <linux/version.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/uaccess.h>
+#include <linux/vmalloc.h>
+#include <linux/poll.h>
+#include <linux/kprobes.h>
+#include <linux/interrupt.h>
+#include <linux/debugfs.h>
+#include <linux/slab.h>
+#include <linux/ring_buffer.h>
+#include <asm/gtp.h>
+
+/* If define GTP_FRAME_SIMPLE, KGTP will use simple frame allocer replace
+ ring buffer. */
+/* #define GTP_FRAME_SIMPLE */
+
+/* If define GTP_CLOCK_CYCLE, $clock will return rdtscll. */
+/* #define GTP_CLOCK_CYCLE */
+
+#ifdef GTP_FRAME_SIMPLE
+/* This define is for simple frame alloc record, then we can get how many
+ memory are weste by FRAME_ALIGN. */
+/* #define FRAME_ALLOC_RECORD */
+#endif
+
+#define KERN_NULL
+
+#ifdef GTPDEBUG
+#define GTP_DEBUG KERN_WARNING
+#endif
+
+#define GTP_RW_MAX 16384
+
+#define FID_TYPE unsigned int
+#define FID_SIZE sizeof(FID_TYPE)
+#define FID(x) (*((FID_TYPE *)x))
+#define FID_HEAD 0
+#define FID_REG 1
+#define FID_MEM 2
+#define FID_VAR 3
+#define FID_END 4
+
+/* GTP_FRAME_SIZE must align with FRAME_ALIGN_SIZE if use GTP_FRAME_SIMPLE. */
+#define GTP_FRAME_SIZE 5242880
+#ifdef GTP_FRAME_SIMPLE
+#define FRAME_ALIGN_SIZE sizeof(unsigned int)
+#define FRAME_ALIGN(x) ((x + FRAME_ALIGN_SIZE - 1) \
+ & (~(FRAME_ALIGN_SIZE - 1)))
+#define GTP_FRAME_HEAD_SIZE (FID_SIZE + sizeof(char *) + sizeof(ULONGEST))
+#define GTP_FRAME_REG_SIZE (FID_SIZE + sizeof(char *) \
+ + sizeof(struct pt_regs))
+#define GTP_FRAME_MEM_SIZE (FID_SIZE + sizeof(char *) \
+ + sizeof(struct gtp_frame_mem))
+#define GTP_FRAME_VAR_SIZE (FID_SIZE + sizeof(char *) \
+ + sizeof(struct gtp_frame_var))
+#else
+#define GTP_FRAME_HEAD_SIZE (FID_SIZE + sizeof(ULONGEST))
+#define GTP_FRAME_REG_SIZE (FID_SIZE + sizeof(struct pt_regs))
+#define GTP_FRAME_MEM_SIZE (FID_SIZE + sizeof(struct gtp_frame_mem))
+#define GTP_FRAME_VAR_SIZE (FID_SIZE + sizeof(struct gtp_frame_var))
+#endif
+#include <linux/perf_event.h>
+
+#define TOHEX(h) ((h) > 9 ? (h) + 'a' - 10 : (h) + '0')
+
+struct action_agent_exp {
+ unsigned int size;
+ uint8_t *buf;
+ int need_var_lock;
+};
+
+struct action_m {
+ int regnum;
+ CORE_ADDR offset;
+ size_t size;
+};
+
+struct action {
+ struct action *next;
+ unsigned char type;
+ char *src;
+ union {
+ ULONGEST reg_mask;
+ struct action_agent_exp exp;
+ struct action_m m;
+ } u;
+};
+
+struct gtpsrc {
+ struct gtpsrc *next;
+ char *src;
+};
+
+enum gtp_stop_type {
+ gtp_stop_normal = 0,
+ gtp_stop_frame_full,
+ gtp_stop_efault,
+ gtp_stop_access_wrong_reg,
+ gtp_stop_agent_expr_code_error,
+ gtp_stop_agent_expr_stack_overflow,
+};
+
+struct gtp_entry {
+ int kpreg;
+ int no_self_trace;
+ int nopass;
+ int have_printk;
+ ULONGEST num;
+ struct action *cond;
+ struct action *action_list;
+ atomic_t current_pass;
+ struct gtpsrc *printk_str;
+ enum gtp_stop_type reason;
+ struct tasklet_struct tasklet;
+ struct work_struct work;
+ struct gtp_entry *next;
+ struct kretprobe kp;
+ int disable;
+ int is_kretprobe;
+ ULONGEST addr;
+ ULONGEST pass;
+ struct gtpsrc *src;
+ ULONGEST step;
+};
+
+#ifdef CONFIG_PERF_EVENTS
+struct pe_tv_s {
+ struct pe_tv_s *pc_next;
+ int en;
+ struct perf_event *event;
+ int cpu;
+ u64 val;
+ u64 enabled; /* The perf inside timer */
+ u64 running; /* The perf inside timer */
+ char *name;
+ struct perf_event_attr attr;
+};
+#endif
+
+enum pe_tv_id {
+ pe_tv_unknown = 0,
+ pe_tv_cpu,
+ pe_tv_type,
+ pe_tv_config,
+ pe_tv_en,
+ pe_tv_val,
+ pe_tv_enabled,
+ pe_tv_running,
+};
+
+struct gtp_var {
+ struct gtp_var *next;
+ unsigned int num;
+ uint64_t val;
+ char *src;
+ struct gtp_var **per_cpu;
+#ifdef CONFIG_PERF_EVENTS
+ enum pe_tv_id ptid;
+ struct pe_tv_s *pts;
+#endif
+};
+
+struct gtp_frame_mem {
+ CORE_ADDR addr;
+ size_t size;
+};
+
+struct gtp_frame_var {
+ unsigned int num;
+ uint64_t val;
+};
+
+struct gtpro_entry {
+ struct gtpro_entry *next;
+ CORE_ADDR start;
+ CORE_ADDR end;
+};
+
+static pid_t gtp_gtp_pid;
+static unsigned int gtp_gtp_pid_count;
+static pid_t gtp_gtpframe_pid;
+static unsigned int gtp_gtpframe_pid_count;
+#ifndef GTP_FRAME_SIMPLE
+static pid_t gtp_gtpframe_pipe_pid;
+static unsigned int gtp_gtpframe_pipe_pid_count;
+#endif
+
+static struct gtp_entry *gtp_list;
+static struct gtp_entry *current_gtp;
+static struct action *current_gtp_action;
+static struct gtpsrc *current_gtp_src;
+
+static struct workqueue_struct *gtp_wq;
+
+static char gtp_read_ack;
+static char *gtp_rw_buf;
+static char *gtp_rw_bufp;
+static size_t gtp_rw_size;
+
+static int gtp_start;
+
+static int gtp_disconnected_tracing;
+static int gtp_circular;
+
+static int gtp_cpu_number;
+
+static DEFINE_SPINLOCK(gtp_var_lock);
+static struct gtp_var *gtp_var_list;
+static unsigned int gtp_var_head;
+static unsigned int gtp_var_tail;
+static struct gtp_var **gtp_var_array;
+static struct gtp_var *current_gtp_var;
+
+enum {
+ GTP_VAR_SPECIAL_MIN = 1,
+ GTP_VAR_CURRENT_TASK_ID = GTP_VAR_SPECIAL_MIN,
+ GTP_VAR_CURRENT_THREAD_INFO_ID,
+ GTP_VAR_CLOCK_ID,
+ GTP_VAR_COOKED_CLOCK_ID,
+#ifdef CONFIG_X86
+ GTP_VAR_RDTSC_ID,
+ GTP_VAR_COOKED_RDTSC_ID,
+#endif
+ GTP_VAR_CPU_ID,
+ GTP_VAR_PRINTK_TMP_ID,
+ GTP_VAR_PRINTK_LEVEL_ID,
+ GTP_VAR_PRINTK_FORMAT_ID,
+ GTP_VAR_DUMP_STACK_ID,
+ GTP_VAR_NO_SELF_TRACE_ID,
+ GTP_VAR_CPU_NUMBER_ID,
+ GTP_VAR_PC_PE_EN_ID,
+ GTP_VAR_KRET_ID,
+ GTP_VAR_XTIME_SEC_ID,
+ GTP_VAR_XTIME_NSEC_ID,
+ GTP_VAR_LAST_ERRNO_ID,
+ GTP_VAR_HARDIRQ_COUNT_ID,
+ GTP_VAR_SOFTIRQ_COUNT_ID,
+ GTP_VAR_IRQ_COUNT_ID,
+ GTP_VAR_SPECIAL_MAX = GTP_VAR_IRQ_COUNT_ID,
+};
+#define PREV_VAR NULL
+static struct gtp_var gtp_var_current_task = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_CURRENT_TASK_ID,
+ .src = "0:1:63757272656e745f7461736b",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_current_task)
+static struct gtp_var gtp_var_current_thread_info = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_CURRENT_THREAD_INFO_ID,
+ .src = "0:1:63757272656e745f7468726561645f696e666f",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_current_thread_info)
+static struct gtp_var gtp_var_clock = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_CLOCK_ID,
+ .src = "0:1:636c6f636b",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_clock)
+static struct gtp_var gtp_var_cooked_clock = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_COOKED_CLOCK_ID,
+ .src = "0:1:636f6f6b65645f636c6f636b",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_cooked_clock)
+#ifdef CONFIG_X86
+static struct gtp_var gtp_var_rdtsc = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_RDTSC_ID,
+ .src = "0:1:7264747363",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_rdtsc)
+static struct gtp_var gtp_var_cooked_rdtsc = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_COOKED_RDTSC_ID,
+ .src = "0:1:636f6f6b65645f7264747363",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_cooked_rdtsc)
+#endif
+static struct gtp_var gtp_var_cpu_id = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_CPU_ID,
+ .src = "0:1:6370755f6964",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_cpu_id)
+static struct gtp_var gtp_var_printk_tmp = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_PRINTK_TMP_ID,
+ .src = "0:1:7072696e746b5f746d70",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_printk_tmp)
+static struct gtp_var gtp_var_printk_level = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_PRINTK_LEVEL_ID,
+ .src = "8:1:7072696e746b5f6c6576656c",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_printk_level)
+static struct gtp_var gtp_var_printk_format = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_PRINTK_FORMAT_ID,
+ .src = "0:1:7072696e746b5f666f726d6174",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_printk_format)
+static struct gtp_var gtp_var_dump_stack = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_DUMP_STACK_ID,
+ .src = "0:1:64756d705f737461636b",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_dump_stack)
+static struct gtp_var gtp_var_no_self_trace = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_NO_SELF_TRACE_ID,
+ .src = "0:1:6e6f5f73656c665f7472616365",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_no_self_trace)
+static struct gtp_var gtp_var_cpu_number = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_CPU_NUMBER_ID,
+ .src = "0:1:6370755f6e756d626572",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_cpu_number)
+static struct gtp_var gtp_var_pc_pe_en = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_PC_PE_EN_ID,
+ .src = "0:1:70635f70655f656e",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_pc_pe_en)
+static struct gtp_var gtp_var_kret = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_KRET_ID,
+ .src = "0:1:6b726574",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_kret)
+static struct gtp_var gtp_var_xtime_sec = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_XTIME_SEC_ID,
+ .src = "0:1:7874696d655f736563",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_xtime_sec)
+static struct gtp_var gtp_var_xtime_nsec = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_XTIME_NSEC_ID,
+ .src = "0:1:7874696d655f6e736563",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_xtime_nsec)
+static struct gtp_var gtp_var_last_errno = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_LAST_ERRNO_ID,
+ .src = "0:1:6c6173745f6572726e6f",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_last_errno)
+static struct gtp_var gtp_var_hardirq_count = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_HARDIRQ_COUNT_ID,
+ .src = "0:1:686172646972715f636f756e74",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_hardirq_count)
+static struct gtp_var gtp_var_softirq_count = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_SOFTIRQ_COUNT_ID,
+ .src = "0:1:736f66746972715f636f756e74",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define PREV_VAR (>p_var_softirq_count)
+static struct gtp_var gtp_var_irq_count = {
+ .next = PREV_VAR,
+ .num = GTP_VAR_IRQ_COUNT_ID,
+ .src = "0:1:6972715f636f756e74",
+ .per_cpu = NULL,
+};
+#undef PREV_VAR
+#define GTP_VAR_LIST_FIRST (>p_var_irq_count)
+
+#define GTP_VAR_IS_SPECIAL(x) ((x) >= GTP_VAR_SPECIAL_MIN \
+ && (x) <= GTP_VAR_SPECIAL_MAX)
+#define GTP_VAR_NOT_GETV(x) ((x) == GTP_VAR_PRINTK_LEVEL_ID \
+ || (x) == GTP_VAR_PRINTK_FORMAT_ID \
+ || (x) == GTP_VAR_PC_PE_EN_ID \
+ || (x) == GTP_VAR_KRET_ID)
+#define GTP_VAR_NOT_SETV(x) (((x) >= GTP_VAR_CURRENT_TASK_ID \
+ && (x) <= GTP_VAR_CPU_ID) \
+ || (x) == GTP_VAR_DUMP_STACK_ID \
+ || (x) == GTP_VAR_CPU_NUMBER_ID \
+ || (x) == GTP_VAR_KRET_ID)
+#define GTP_VAR_NOT_TRACEV(x) (((x) >= GTP_VAR_PRINTK_LEVEL_ID \
+ && (x) <= GTP_VAR_PRINTK_FORMAT_ID) \
+ || (x) == GTP_VAR_PC_PE_EN_ID \
+ || (x) == GTP_VAR_KRET_ID)
+#define GTP_VAR_AUTO_TRACEV(x) ((x) == GTP_VAR_CLOCK_ID \
+ || (x) == GTP_VAR_CPU_ID)
+
+static int gtp_frame_current_num;
+static ULONGEST gtp_frame_current_tpe;
+static atomic_t gtp_frame_create;
+static char *gtp_frame_file;
+static size_t gtp_frame_file_size;
+static DECLARE_WAIT_QUEUE_HEAD(gtpframe_wq);
+#ifdef GTP_FRAME_SIMPLE
+static DEFINE_SPINLOCK(gtp_frame_lock);
+static char *gtp_frame;
+static char *gtp_frame_r_start;
+static char *gtp_frame_w_start;
+static char *gtp_frame_end;
+static int gtp_frame_is_circular;
+static char *gtp_frame_current;
+#else
+static struct ring_buffer *gtp_frame;
+static struct ring_buffer_iter *gtp_frame_iter[NR_CPUS];
+static int gtp_frame_current_cpu;
+static u64 gtp_frame_current_clock;
+
+static DECLARE_WAIT_QUEUE_HEAD(gtpframe_pipe_wq);
+static atomic_t gtpframe_pipe_wq_v;
+static struct tasklet_struct gtpframe_pipe_wq_tasklet;
+#endif
+
+static struct gtpro_entry *gtpro_list;
+
+#define GTP_PRINTF_MAX 256
+static DEFINE_PER_CPU(char[GTP_PRINTF_MAX], gtp_printf);
+
+static DEFINE_PER_CPU(u64, rdtsc_current);
+static DEFINE_PER_CPU(u64, local_clock_current);
+static DEFINE_PER_CPU(u64, rdtsc_offset);
+static DEFINE_PER_CPU(u64, local_clock_offset);
+
+#define GTP_LOCAL_CLOCK local_clock()
+
+#ifdef CONFIG_X86
+static ULONGEST
+gtp_action_reg_read(struct pt_regs *regs, struct gtp_entry *tpe, int num)
+{
+ ULONGEST ret;
+
+ switch (num) {
+#ifdef CONFIG_X86_32
+ case 0:
+ ret = regs->ax;
+ break;
+ case 1:
+ ret = regs->cx;
+ break;
+ case 2:
+ ret = regs->dx;
+ break;
+ case 3:
+ ret = regs->bx;
+ break;
+ case 4:
+ ret = (ULONGEST)(CORE_ADDR)®s->sp;
+ break;
+ case 5:
+ ret = regs->bp;
+ break;
+ case 6:
+ ret = regs->si;
+ break;
+ case 7:
+ ret = regs->di;
+ break;
+ case 8:
+ ret = regs->ip - 1;
+ break;
+ case 9:
+ ret = regs->flags;
+ break;
+ case 10:
+ ret = regs->cs;
+ break;
+ case 11:
+ ret = regs->ss;
+ break;
+ case 12:
+ ret = regs->ds;
+ break;
+ case 13:
+ ret = regs->es;
+ break;
+ case 14:
+ ret = regs->fs;
+ break;
+ case 15:
+ ret = regs->gs;
+ break;
+#else
+ case 0:
+ ret = regs->ax;
+ break;
+ case 1:
+ ret = regs->bx;
+ break;
+ case 2:
+ ret = regs->cx;
+ break;
+ case 3:
+ ret = regs->dx;
+ break;
+ case 4:
+ ret = regs->si;
+ break;
+ case 5:
+ ret = regs->di;
+ break;
+ case 6:
+ ret = regs->bp;
+ break;
+ case 7:
+ ret = regs->sp;
+ break;
+ case 8:
+ ret = regs->r8;
+ break;
+ case 9:
+ ret = regs->r9;
+ break;
+ case 10:
+ ret = regs->r10;
+ break;
+ case 11:
+ ret = regs->r11;
+ break;
+ case 12:
+ ret = regs->r12;
+ break;
+ case 13:
+ ret = regs->r13;
+ break;
+ case 14:
+ ret = regs->r14;
+ break;
+ case 15:
+ ret = regs->r15;
+ break;
+ case 16:
+ ret = regs->ip - 1;
+ break;
+ case 17:
+ ret = regs->flags;
+ break;
+ case 18:
+ ret = regs->cs;
+ break;
+ case 19:
+ ret = regs->ss;
+ break;
+#endif
+ default:
+ ret = 0;
+ tpe->reason = gtp_stop_access_wrong_reg;
+ break;
+ }
+
+ return ret;
+}
+
+static void
+gtp_regs2ascii(struct pt_regs *regs, char *buf)
+{
+#ifdef CONFIG_X86_32
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_regs2ascii: ax = 0x%x\n",
+ (unsigned int) regs->ax);
+ printk(GTP_DEBUG "gtp_regs2ascii: cx = 0x%x\n",
+ (unsigned int) regs->cx);
+ printk(GTP_DEBUG "gtp_regs2ascii: dx = 0x%x\n",
+ (unsigned int) regs->dx);
+ printk(GTP_DEBUG "gtp_regs2ascii: bx = 0x%x\n",
+ (unsigned int) regs->bx);
+ printk(GTP_DEBUG "gtp_regs2ascii: sp = 0x%x\n",
+ (unsigned int) regs->sp);
+ printk(GTP_DEBUG "gtp_regs2ascii: bp = 0x%x\n",
+ (unsigned int) regs->bp);
+ printk(GTP_DEBUG "gtp_regs2ascii: si = 0x%x\n",
+ (unsigned int) regs->si);
+ printk(GTP_DEBUG "gtp_regs2ascii: di = 0x%x\n",
+ (unsigned int) regs->di);
+ printk(GTP_DEBUG "gtp_regs2ascii: ip = 0x%x\n",
+ (unsigned int) regs->ip);
+ printk(GTP_DEBUG "gtp_regs2ascii: flags = 0x%x\n",
+ (unsigned int) regs->flags);
+ printk(GTP_DEBUG "gtp_regs2ascii: cs = 0x%x\n",
+ (unsigned int) regs->cs);
+ printk(GTP_DEBUG "gtp_regs2ascii: ss = 0x%x\n",
+ (unsigned int) regs->ss);
+ printk(GTP_DEBUG "gtp_regs2ascii: ds = 0x%x\n",
+ (unsigned int) regs->ds);
+ printk(GTP_DEBUG "gtp_regs2ascii: es = 0x%x\n",
+ (unsigned int) regs->es);
+ printk(GTP_DEBUG "gtp_regs2ascii: fs = 0x%x\n",
+ (unsigned int) regs->fs);
+ printk(GTP_DEBUG "gtp_regs2ascii: gs = 0x%x\n",
+ (unsigned int) regs->gs);
+#endif
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->ax));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->cx));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->dx));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->bx));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->sp));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->bp));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->si));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->di));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->ip));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->flags));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->cs));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->ss));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->ds));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->es));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->fs));
+ buf += 8;
+ sprintf(buf, "%08x", (unsigned int) swab32(regs->gs));
+ buf += 8;
+#else
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_regs2ascii: ax = 0x%lx\n", regs->ax);
+ printk(GTP_DEBUG "gtp_regs2ascii: bx = 0x%lx\n", regs->bx);
+ printk(GTP_DEBUG "gtp_regs2ascii: cx = 0x%lx\n", regs->cx);
+ printk(GTP_DEBUG "gtp_regs2ascii: dx = 0x%lx\n", regs->dx);
+ printk(GTP_DEBUG "gtp_regs2ascii: si = 0x%lx\n", regs->si);
+ printk(GTP_DEBUG "gtp_regs2ascii: di = 0x%lx\n", regs->di);
+ printk(GTP_DEBUG "gtp_regs2ascii: bp = 0x%lx\n", regs->bp);
+ printk(GTP_DEBUG "gtp_regs2ascii: sp = 0x%lx\n", regs->sp);
+ printk(GTP_DEBUG "gtp_regs2ascii: r8 = 0x%lx\n", regs->r8);
+ printk(GTP_DEBUG "gtp_regs2ascii: r9 = 0x%lx\n", regs->r9);
+ printk(GTP_DEBUG "gtp_regs2ascii: r10 = 0x%lx\n", regs->r10);
+ printk(GTP_DEBUG "gtp_regs2ascii: r11 = 0x%lx\n", regs->r11);
+ printk(GTP_DEBUG "gtp_regs2ascii: r12 = 0x%lx\n", regs->r12);
+ printk(GTP_DEBUG "gtp_regs2ascii: r13 = 0x%lx\n", regs->r13);
+ printk(GTP_DEBUG "gtp_regs2ascii: r14 = 0x%lx\n", regs->r14);
+ printk(GTP_DEBUG "gtp_regs2ascii: r15 = 0x%lx\n", regs->r15);
+ printk(GTP_DEBUG "gtp_regs2ascii: ip = 0x%lx\n", regs->ip);
+ printk(GTP_DEBUG "gtp_regs2ascii: flags = 0x%lx\n", regs->flags);
+ printk(GTP_DEBUG "gtp_regs2ascii: cs = 0x%lx\n", regs->cs);
+ printk(GTP_DEBUG "gtp_regs2ascii: ss = 0x%lx\n", regs->ss);
+#endif
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->ax));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->bx));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->cx));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->dx));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->si));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->di));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->bp));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->sp));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->r8));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->r9));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->r10));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->r11));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->r12));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->r13));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->r14));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->r15));
+ buf += 16;
+ sprintf(buf, "%016lx", (unsigned long) swab64(regs->ip));
+ buf += 16;
+ sprintf(buf, "%08x",
+ (unsigned int) swab32((unsigned int)regs->flags));
+ buf += 8;
+ sprintf(buf, "%08x",
+ (unsigned int) swab32((unsigned int)regs->cs));
+ buf += 8;
+ sprintf(buf, "%08x",
+ (unsigned int) swab32((unsigned int)regs->ss));
+ buf += 8;
+#endif
+}
+
+static void
+gtp_regs2bin(struct pt_regs *regs, char *buf)
+{
+#ifdef CONFIG_X86_32
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_regs2ascii: ax = 0x%x\n",
+ (unsigned int) regs->ax);
+ printk(GTP_DEBUG "gtp_regs2ascii: cx = 0x%x\n",
+ (unsigned int) regs->cx);
+ printk(GTP_DEBUG "gtp_regs2ascii: dx = 0x%x\n",
+ (unsigned int) regs->dx);
+ printk(GTP_DEBUG "gtp_regs2ascii: bx = 0x%x\n",
+ (unsigned int) regs->bx);
+ printk(GTP_DEBUG "gtp_regs2ascii: sp = 0x%x\n",
+ (unsigned int) regs->sp);
+ printk(GTP_DEBUG "gtp_regs2ascii: bp = 0x%x\n",
+ (unsigned int) regs->bp);
+ printk(GTP_DEBUG "gtp_regs2ascii: si = 0x%x\n",
+ (unsigned int) regs->si);
+ printk(GTP_DEBUG "gtp_regs2ascii: di = 0x%x\n",
+ (unsigned int) regs->di);
+ printk(GTP_DEBUG "gtp_regs2ascii: ip = 0x%x\n",
+ (unsigned int) regs->ip);
+ printk(GTP_DEBUG "gtp_regs2ascii: flags = 0x%x\n",
+ (unsigned int) regs->flags);
+ printk(GTP_DEBUG "gtp_regs2ascii: cs = 0x%x\n",
+ (unsigned int) regs->cs);
+ printk(GTP_DEBUG "gtp_regs2ascii: ss = 0x%x\n",
+ (unsigned int) regs->ss);
+ printk(GTP_DEBUG "gtp_regs2ascii: ds = 0x%x\n",
+ (unsigned int) regs->ds);
+ printk(GTP_DEBUG "gtp_regs2ascii: es = 0x%x\n",
+ (unsigned int) regs->es);
+ printk(GTP_DEBUG "gtp_regs2ascii: fs = 0x%x\n",
+ (unsigned int) regs->fs);
+ printk(GTP_DEBUG "gtp_regs2ascii: gs = 0x%x\n",
+ (unsigned int) regs->gs);
+#endif
+ memcpy(buf, ®s->ax, 4);
+ buf += 4;
+ memcpy(buf, ®s->cx, 4);
+ buf += 4;
+ memcpy(buf, ®s->dx, 4);
+ buf += 4;
+ memcpy(buf, ®s->bx, 4);
+ buf += 4;
+ memcpy(buf, ®s->sp, 4);
+ buf += 4;
+ memcpy(buf, ®s->bp, 4);
+ buf += 4;
+ memcpy(buf, ®s->si, 4);
+ buf += 4;
+ memcpy(buf, ®s->di, 4);
+ buf += 4;
+ memcpy(buf, ®s->ip, 4);
+ buf += 4;
+ memcpy(buf, ®s->flags, 4);
+ buf += 4;
+ memcpy(buf, ®s->cs, 4);
+ buf += 4;
+ memcpy(buf, ®s->ss, 4);
+ buf += 4;
+ memcpy(buf, ®s->ds, 4);
+ buf += 4;
+ memcpy(buf, ®s->es, 4);
+ buf += 4;
+ memcpy(buf, ®s->fs, 4);
+ buf += 4;
+ memcpy(buf, ®s->gs, 4);
+ buf += 4;
+#else
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_regs2ascii: ax = 0x%lx\n", regs->ax);
+ printk(GTP_DEBUG "gtp_regs2ascii: bx = 0x%lx\n", regs->bx);
+ printk(GTP_DEBUG "gtp_regs2ascii: cx = 0x%lx\n", regs->cx);
+ printk(GTP_DEBUG "gtp_regs2ascii: dx = 0x%lx\n", regs->dx);
+ printk(GTP_DEBUG "gtp_regs2ascii: si = 0x%lx\n", regs->si);
+ printk(GTP_DEBUG "gtp_regs2ascii: di = 0x%lx\n", regs->di);
+ printk(GTP_DEBUG "gtp_regs2ascii: bp = 0x%lx\n", regs->bp);
+ printk(GTP_DEBUG "gtp_regs2ascii: sp = 0x%lx\n", regs->sp);
+ printk(GTP_DEBUG "gtp_regs2ascii: r8 = 0x%lx\n", regs->r8);
+ printk(GTP_DEBUG "gtp_regs2ascii: r9 = 0x%lx\n", regs->r9);
+ printk(GTP_DEBUG "gtp_regs2ascii: r10 = 0x%lx\n", regs->r10);
+ printk(GTP_DEBUG "gtp_regs2ascii: r11 = 0x%lx\n", regs->r11);
+ printk(GTP_DEBUG "gtp_regs2ascii: r12 = 0x%lx\n", regs->r12);
+ printk(GTP_DEBUG "gtp_regs2ascii: r13 = 0x%lx\n", regs->r13);
+ printk(GTP_DEBUG "gtp_regs2ascii: r14 = 0x%lx\n", regs->r14);
+ printk(GTP_DEBUG "gtp_regs2ascii: r15 = 0x%lx\n", regs->r15);
+ printk(GTP_DEBUG "gtp_regs2ascii: ip = 0x%lx\n", regs->ip);
+ printk(GTP_DEBUG "gtp_regs2ascii: flags = 0x%lx\n", regs->flags);
+ printk(GTP_DEBUG "gtp_regs2ascii: cs = 0x%lx\n", regs->cs);
+ printk(GTP_DEBUG "gtp_regs2ascii: ss = 0x%lx\n", regs->ss);
+#endif
+ memcpy(buf, ®s->ax, 8);
+ buf += 8;
+ memcpy(buf, ®s->bx, 8);
+ buf += 8;
+ memcpy(buf, ®s->cx, 8);
+ buf += 8;
+ memcpy(buf, ®s->dx, 8);
+ buf += 8;
+ memcpy(buf, ®s->si, 8);
+ buf += 8;
+ memcpy(buf, ®s->di, 8);
+ buf += 8;
+ memcpy(buf, ®s->bp, 8);
+ buf += 8;
+ memcpy(buf, ®s->sp, 8);
+ buf += 8;
+ memcpy(buf, ®s->r8, 8);
+ buf += 8;
+ memcpy(buf, ®s->r9, 8);
+ buf += 8;
+ memcpy(buf, ®s->r10, 8);
+ buf += 8;
+ memcpy(buf, ®s->r11, 8);
+ buf += 8;
+ memcpy(buf, ®s->r12, 8);
+ buf += 8;
+ memcpy(buf, ®s->r13, 8);
+ buf += 8;
+ memcpy(buf, ®s->r14, 8);
+ buf += 8;
+ memcpy(buf, ®s->r15, 8);
+ buf += 8;
+ memcpy(buf, ®s->ip, 8);
+ buf += 8;
+ memcpy(buf, ®s->flags, 4);
+ buf += 4;
+ memcpy(buf, ®s->cs, 4);
+ buf += 4;
+ memcpy(buf, ®s->ss, 4);
+ buf += 4;
+#endif
+}
+#endif
+
+#ifdef CONFIG_MIPS
+static ULONGEST
+gtp_action_reg_read(struct pt_regs *regs, struct gtp_entry *tpe, int num)
+{
+ ULONGEST ret;
+
+ if (num > 90) {
+ /* GDB convert the reg number to a GDB
+ [1 * gdbarch_num_regs .. 2 * gdbarch_num_regs) REGNUM
+ in function mips_dwarf_dwarf2_ecoff_reg_to_regnum. */
+ num -= 90;
+ }
+
+ if (num >= 0 && num <= 31) {
+ ret = regs->regs[num];
+ } else {
+ switch (num) {
+ case 32:
+ ret = regs->cp0_status;
+ break;
+ case 33:
+ ret = regs->lo;
+ break;
+ case 34:
+ ret = regs->hi;
+ break;
+ case 35:
+ ret = regs->cp0_badvaddr;
+ break;
+ case 36:
+ ret = regs->cp0_cause;
+ break;
+ case 37:
+ ret = regs->cp0_epc;
+ break;
+ default:
+ ret = 0;
+ tpe->reason = gtp_stop_access_wrong_reg;
+ break;
+ }
+ }
+
+ return ret;
+};
+
+static void
+gtp_regs2ascii(struct pt_regs *regs, char *buf)
+{
+#ifdef GTP_DEBUG
+ {
+ int i;
+
+ for (i = 0; i < 32; i++)
+ printk(GTP_DEBUG "gtp_gdbrsp_g: r%d = 0x%lx\n", i,
+ regs->regs[i]);
+ }
+ printk(GTP_DEBUG "gtp_gdbrsp_g: status = 0x%lx\n",
+ regs->cp0_status);
+ printk(GTP_DEBUG "gtp_gdbrsp_g: lo = 0x%lx\n", regs->lo);
+ printk(GTP_DEBUG "gtp_gdbrsp_g: hi = 0x%lx\n", regs->hi);
+ printk(GTP_DEBUG "gtp_gdbrsp_g: badvaddr = 0x%lx\n",
+ regs->cp0_badvaddr);
+ printk(GTP_DEBUG "gtp_gdbrsp_g: cause = 0x%lx\n", regs->cp0_cause);
+ printk(GTP_DEBUG "gtp_gdbrsp_g: pc = 0x%lx\n", regs->cp0_epc);
+#endif
+
+#ifdef CONFIG_32BIT
+#define OUTFORMAT "%08lx"
+#define REGSIZE 8
+#ifdef __LITTLE_ENDIAN
+#define SWAB(a) swab32(a)
+#else
+#define SWAB(a) (a)
+#endif
+#else
+#define OUTFORMAT "%016lx"
+#define REGSIZE 16
+#ifdef __LITTLE_ENDIAN
+#define SWAB(a) swab64(a)
+#else
+#define SWAB(a) (a)
+#endif
+#endif
+ {
+ int i;
+
+ for (i = 0; i < 32; i++) {
+ sprintf(buf, OUTFORMAT,
+ (unsigned long) SWAB(regs->regs[i]));
+ buf += REGSIZE;
+ }
+ }
+
+ sprintf(buf, OUTFORMAT,
+ (unsigned long) SWAB(regs->cp0_status));
+ buf += REGSIZE;
+ sprintf(buf, OUTFORMAT,
+ (unsigned long) SWAB(regs->lo));
+ buf += REGSIZE;
+ sprintf(buf, OUTFORMAT,
+ (unsigned long) SWAB(regs->hi));
+ buf += REGSIZE;
+ sprintf(buf, OUTFORMAT,
+ (unsigned long) SWAB(regs->cp0_badvaddr));
+ buf += REGSIZE;
+ sprintf(buf, OUTFORMAT,
+ (unsigned long) SWAB(regs->cp0_cause));
+ buf += REGSIZE;
+ sprintf(buf, OUTFORMAT,
+ (unsigned long) SWAB(regs->cp0_epc));
+ buf += REGSIZE;
+#undef OUTFORMAT
+#undef REGSIZE
+#undef SWAB
+}
+
+static void
+gtp_regs2bin(struct pt_regs *regs, char *buf)
+{
+#ifdef GTP_DEBUG
+ {
+ int i;
+
+ for (i = 0; i < 32; i++)
+ printk(GTP_DEBUG "gtp_gdbrsp_g: r%d = 0x%lx\n", i,
+ regs->regs[i]);
+ }
+ printk(GTP_DEBUG "gtp_gdbrsp_g: status = 0x%lx\n",
+ regs->cp0_status);
+ printk(GTP_DEBUG "gtp_gdbrsp_g: lo = 0x%lx\n", regs->lo);
+ printk(GTP_DEBUG "gtp_gdbrsp_g: hi = 0x%lx\n", regs->hi);
+ printk(GTP_DEBUG "gtp_gdbrsp_g: badvaddr = 0x%lx\n",
+ regs->cp0_badvaddr);
+ printk(GTP_DEBUG "gtp_gdbrsp_g: cause = 0x%lx\n", regs->cp0_cause);
+ printk(GTP_DEBUG "gtp_gdbrsp_g: pc = 0x%lx\n", regs->cp0_epc);
+#endif
+
+#ifdef CONFIG_32BIT
+#define REGSIZE 4
+#else
+#define REGSIZE 8
+#endif
+ {
+ int i;
+
+ for (i = 0; i < 32; i++) {
+ memcpy(buf, ®s->regs[i], REGSIZE);
+ buf += REGSIZE;
+ }
+ }
+ memcpy(buf, ®s->cp0_status, REGSIZE);
+ buf += REGSIZE;
+ memcpy(buf, ®s->lo, REGSIZE);
+ buf += REGSIZE;
+ memcpy(buf, ®s->hi, REGSIZE);
+ buf += REGSIZE;
+ memcpy(buf, ®s->cp0_badvaddr, REGSIZE);
+ buf += REGSIZE;
+ memcpy(buf, ®s->cp0_cause, REGSIZE);
+ buf += REGSIZE;
+ memcpy(buf, ®s->cp0_epc, REGSIZE);
+ buf += REGSIZE;
+#undef REGSIZE
+}
+#endif
+
+#ifdef CONFIG_ARM
+static ULONGEST
+gtp_action_reg_read(struct pt_regs *regs, struct gtp_entry *tpe, int num)
+{
+ if (num >= 0 && num < 16)
+ return regs->uregs[num];
+ else if (num == 25)
+ return regs->uregs[16];
+
+ tpe->reason = gtp_stop_access_wrong_reg;
+ return 0;
+}
+
+static void
+gtp_regs2ascii(struct pt_regs *regs, char *buf)
+{
+#ifdef __LITTLE_ENDIAN
+#define SWAB(a) swab32(a)
+#else
+#define SWAB(a) (a)
+#endif
+ int i;
+
+ for (i = 0; i < 16; i++) {
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_g: r%d = 0x%lx\n",
+ i, regs->uregs[i]);
+#endif
+ sprintf(buf, "%08lx", (unsigned long) SWAB(regs->uregs[i]));
+ buf += 8;
+ }
+
+ /* f0-f7 fps */
+ memset(buf, '0', 200);
+ buf += 200;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_g: cpsr = 0x%lx\n", regs->uregs[16]);
+#endif
+ sprintf(buf, "%08lx",
+ (unsigned long) SWAB(regs->uregs[16]));
+ buf += 8;
+#undef SWAB
+}
+
+static void
+gtp_regs2bin(struct pt_regs *regs, char *buf)
+{
+ int i;
+
+ for (i = 0; i < 16; i++) {
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_g: r%d = 0x%lx\n",
+ i, regs->uregs[i]);
+#endif
+ memcpy(buf, ®s->uregs[i], 4);
+ buf += 4;
+ }
+
+ /* f0-f7 fps */
+ memset(buf, '\0', 100);
+ buf += 100;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_g: cpsr = 0x%lx\n", regs->uregs[16]);
+#endif
+ memcpy(buf, ®s->uregs[16], 4);
+ buf += 4;
+}
+#endif
+
+#ifdef CONFIG_PERF_EVENTS
+static DEFINE_PER_CPU(int, pc_pe_list_all_disabled);
+static DEFINE_PER_CPU(struct pe_tv_s *, pc_pe_list);
+
+static void
+pc_pe_list_disable(void)
+{
+ struct pe_tv_s *ppl;
+
+ if (__get_cpu_var(pc_pe_list_all_disabled))
+ return;
+
+ for (ppl = __get_cpu_var(pc_pe_list); ppl; ppl = ppl->pc_next) {
+ if (ppl->en)
+ __perf_event_disable(ppl->event);
+ }
+}
+
+static void
+pc_pe_list_enable(void)
+{
+ struct pe_tv_s *ppl;
+
+ if (__get_cpu_var(pc_pe_list_all_disabled))
+ return;
+
+ for (ppl = __get_cpu_var(pc_pe_list); ppl; ppl = ppl->pc_next) {
+ if (ppl->en)
+ __perf_event_enable(ppl->event);
+ }
+}
+
+static void
+gtp_pc_pe_en(int enable)
+{
+ struct pe_tv_s *ppl = __get_cpu_var(pc_pe_list);
+
+ for (ppl = __get_cpu_var(pc_pe_list); ppl; ppl = ppl->pc_next)
+ ppl->en = enable;
+
+ __get_cpu_var(pc_pe_list_all_disabled) = !enable;
+}
+
+static void
+gtp_pe_set_en(struct pe_tv_s *pts, int enable)
+{
+ if (pts->event->cpu != smp_processor_id()) {
+ if (enable)
+ perf_event_enable(pts->event);
+ else
+ perf_event_disable(pts->event);
+ }
+ pts->en = enable;
+}
+#else
+static void
+gtp_pc_pe_en(int enable)
+{
+}
+#endif /* CONFIG_PERF_EVENTS */
+
+#ifdef GTP_FRAME_SIMPLE
+static char *
+gtp_frame_next(char *frame)
+{
+ switch (FID(frame)) {
+ case FID_HEAD:
+ frame += FRAME_ALIGN(GTP_FRAME_HEAD_SIZE);
+ break;
+ case FID_REG:
+ frame += FRAME_ALIGN(GTP_FRAME_REG_SIZE);
+ break;
+ case FID_MEM: {
+ struct gtp_frame_mem *gfm;
+
+ gfm = (struct gtp_frame_mem *) (frame + FID_SIZE
+ + sizeof(char *));
+ frame += FRAME_ALIGN(GTP_FRAME_MEM_SIZE + gfm->size);
+ }
+ break;
+ case FID_VAR:
+ frame += FRAME_ALIGN(GTP_FRAME_VAR_SIZE);
+ break;
+ case FID_END:
+ frame = gtp_frame_end;
+ break;
+ default:
+ return NULL;
+ break;
+ }
+
+ return frame;
+}
+#endif
+
+#ifdef GTP_FRAME_SIMPLE
+#ifdef FRAME_ALLOC_RECORD
+ULONGEST frame_alloc_size;
+ULONGEST frame_alloc_size_hole;
+#endif
+
+static char *
+gtp_frame_alloc(size_t size)
+{
+ char *ret = NULL;
+
+#ifdef FRAME_ALLOC_RECORD
+ frame_alloc_size += size;
+ frame_alloc_size_hole += (FRAME_ALIGN(size) - size);
+#endif
+
+ size = FRAME_ALIGN(size);
+
+ if (size > GTP_FRAME_SIZE)
+ return NULL;
+
+ spin_lock(>p_frame_lock);
+
+ if (gtp_frame_w_start + size > gtp_frame_end) {
+ if (gtp_circular) {
+ gtp_frame_is_circular = 1;
+#ifdef FRAME_ALLOC_RECORD
+ if (gtp_frame_w_start != gtp_frame_end
+ && gtp_frame_end - gtp_frame_w_start < FID_SIZE) {
+ printk(KERN_WARNING "Frame align wrong."
+ "start = %p end = %p\n",
+ gtp_frame_w_start, gtp_frame_end);
+ goto out;
+ }
+#endif
+ if (gtp_frame_w_start != gtp_frame_end)
+ FID(gtp_frame_w_start) = FID_END;
+ gtp_frame_w_start = gtp_frame;
+ gtp_frame_r_start = gtp_frame;
+ } else
+ goto out;
+ }
+
+ if (gtp_frame_is_circular) {
+ while (gtp_frame_w_start <= gtp_frame_r_start
+ && gtp_frame_w_start + size > gtp_frame_r_start) {
+ char *tmp = gtp_frame_next(gtp_frame_r_start);
+ if (tmp == NULL)
+ goto out;
+ if (tmp == gtp_frame_end)
+ gtp_frame_r_start = gtp_frame;
+ else
+ gtp_frame_r_start = tmp;
+ }
+ }
+
+ ret = gtp_frame_w_start;
+ gtp_frame_w_start += size;
+
+out:
+ spin_unlock(>p_frame_lock);
+ return ret;
+}
+#endif
+
+struct gtp_trace_s {
+ struct gtp_entry *tpe;
+ struct pt_regs *regs;
+#ifdef GTP_FRAME_SIMPLE
+ char **next;
+#else
+ char *next;
+#endif
+ struct kretprobe_instance *ri;
+ int *run;
+ struct timespec xtime;
+ ULONGEST printk_tmp;
+ unsigned int printk_level;
+ unsigned int printk_format;
+ struct gtpsrc *printk_str;
+};
+
+#define GTP_PRINTK_FORMAT_A 0
+#define GTP_PRINTK_FORMAT_D 1
+#define GTP_PRINTK_FORMAT_U 2
+#define GTP_PRINTK_FORMAT_X 3
+#define GTP_PRINTK_FORMAT_S 4
+#define GTP_PRINTK_FORMAT_B 5
+
+#ifndef GTP_FRAME_SIMPLE
+#define GTP_FRAME_RINGBUFFER_ALLOC(size) \
+ do { \
+ rbe = ring_buffer_lock_reserve(gtp_frame, size); \
+ if (rbe == NULL) { \
+ gts->tpe->reason = gtp_stop_frame_full; \
+ return -1; \
+ } \
+ tmp = ring_buffer_event_data(rbe); \
+ } while (0)
+#endif
+
+static struct gtp_var *gtp_gtp_var_array_find(unsigned int num);
+static int gtp_collect_var(struct gtp_trace_s *gts,
+ struct gtp_var *tve);
+
+static int
+gtp_action_head(struct gtp_trace_s *gts)
+{
+ char *tmp;
+ ULONGEST *trace_nump;
+#ifndef GTP_FRAME_SIMPLE
+ struct ring_buffer_event *rbe;
+#endif
+
+ /* Get the head. */
+#ifdef GTP_FRAME_SIMPLE
+ tmp = gtp_frame_alloc(GTP_FRAME_HEAD_SIZE);
+ if (!tmp) {
+ gts->tpe->reason = gtp_stop_frame_full;
+ return -1;
+ }
+#else
+ GTP_FRAME_RINGBUFFER_ALLOC(GTP_FRAME_HEAD_SIZE);
+#endif
+
+ FID(tmp) = FID_HEAD;
+ tmp += FID_SIZE;
+
+#ifdef GTP_FRAME_SIMPLE
+ gts->next = (char **)tmp;
+ *(gts->next) = NULL;
+ tmp += sizeof(char *);
+#endif
+
+ trace_nump = (ULONGEST *)tmp;
+ *trace_nump = gts->tpe->num;
+
+#ifndef GTP_FRAME_SIMPLE
+ ring_buffer_unlock_commit(gtp_frame, rbe);
+ gts->next = (char *)1;
+#endif
+
+#ifdef GTP_FRAME_SIMPLE
+ /* Trace $cpu_id and $clock. */
+ {
+ struct gtp_var *tve;
+
+ tve = gtp_gtp_var_array_find(GTP_VAR_CLOCK_ID);
+ if (!tve) {
+ gts->tpe->reason = gtp_stop_agent_expr_code_error;
+ return -1;
+ }
+ if (gtp_collect_var(gts, tve))
+ return -1;
+ tve = gtp_gtp_var_array_find(GTP_VAR_CPU_ID);
+ if (!tve) {
+ gts->tpe->reason = gtp_stop_agent_expr_code_error;
+ return -1;
+ }
+ if (gtp_collect_var(gts, tve))
+ return -1;
+ }
+#endif
+
+ return 0;
+}
+
+static int
+gtp_action_printk(struct gtp_trace_s *gts, ULONGEST addr, size_t size)
+{
+ unsigned int printk_format = gts->printk_format;
+ char *pbuf = __get_cpu_var(gtp_printf);
+
+ if (gts->printk_str == NULL) {
+ gts->tpe->reason = gtp_stop_agent_expr_code_error;
+ printk(KERN_WARNING "gtp_action_printk: id:%d addr:%p "
+ "printk doesn't have var name. Please "
+ "check actions of it.\n",
+ (int)gts->tpe->num, (void *)(CORE_ADDR)gts->tpe->addr);
+ return -1;
+ }
+
+ if (size) {
+ if (size > GTP_PRINTF_MAX - 1)
+ size = GTP_PRINTF_MAX - 1;
+ if (gts->printk_format != GTP_PRINTK_FORMAT_S
+ && gts->printk_format != GTP_PRINTK_FORMAT_B
+ && size > 8)
+ size = 8;
+ if (probe_kernel_read(pbuf, (void *)(CORE_ADDR)addr, size)) {
+ gts->tpe->reason = gtp_stop_efault;
+ printk(KERN_WARNING "gtp_action_printk: id:%d addr:%p "
+ "read %p %u get error.\n",
+ (int)gts->tpe->num,
+ (void *)(CORE_ADDR)gts->tpe->addr,
+ (void *)(CORE_ADDR)addr,
+ (unsigned int)size);
+ return -1;
+ }
+ } else {
+ size = sizeof(ULONGEST);
+ memcpy(pbuf, &addr, sizeof(ULONGEST));
+ }
+
+ if (printk_format == GTP_PRINTK_FORMAT_A) {
+ if (size == 1 || size == 2 || size == 4 || size == 8)
+ printk_format = GTP_PRINTK_FORMAT_U;
+ else
+ printk_format = GTP_PRINTK_FORMAT_B;
+ }
+
+ switch (printk_format) {
+ case GTP_PRINTK_FORMAT_D:
+ switch (size) {
+ case 1:
+ printk(KERN_NULL "<%d>%s%d\n", gts->printk_level,
+ gts->printk_str->src, pbuf[0]);
+ break;
+ case 2:
+ printk(KERN_NULL "<%d>%s%d\n", gts->printk_level,
+ gts->printk_str->src, (int)(*(short *)pbuf));
+ break;
+ case 4:
+ printk(KERN_NULL "<%d>%s%d\n", gts->printk_level,
+ gts->printk_str->src, *(int *)pbuf);
+ break;
+ case 8:
+ printk(KERN_NULL "<%d>%s%lld\n", gts->printk_level,
+ gts->printk_str->src, *(long long *)pbuf);
+ break;
+ default:
+ printk(KERN_WARNING "gtp_action_printk: id:%d addr:%p "
+ "size %d cannot printk.\n",
+ (int)gts->tpe->num,
+ (void *)(CORE_ADDR)gts->tpe->addr,
+ (unsigned int)size);
+ gts->tpe->reason = gtp_stop_agent_expr_code_error;
+ return -1;
+ break;
+ }
+ break;
+ case GTP_PRINTK_FORMAT_U:
+ switch (size) {
+ case 1:
+ printk(KERN_NULL "<%d>%s%u\n", gts->printk_level,
+ gts->printk_str->src, pbuf[0]);
+ break;
+ case 2:
+ printk(KERN_NULL "<%d>%s%u\n", gts->printk_level,
+ gts->printk_str->src, (int)(*(short *)pbuf));
+ break;
+ case 4:
+ printk(KERN_NULL "<%d>%s%u\n", gts->printk_level,
+ gts->printk_str->src, *(int *)pbuf);
+ break;
+ case 8:
+ printk(KERN_NULL "<%d>%s%llu\n", gts->printk_level,
+ gts->printk_str->src, *(long long *)pbuf);
+ break;
+ default:
+ printk(KERN_WARNING "gtp_action_printk: id:%d addr:%p"
+ "size %d cannot printk.\n",
+ (int)gts->tpe->num,
+ (void *)(CORE_ADDR)gts->tpe->addr,
+ (unsigned int)size);
+ gts->tpe->reason = gtp_stop_agent_expr_code_error;
+ return -1;
+ break;
+ }
+ break;
+ case GTP_PRINTK_FORMAT_X:
+ switch (size) {
+ case 1:
+ printk(KERN_NULL "<%d>%s0x%x\n", gts->printk_level,
+ gts->printk_str->src, pbuf[0]);
+ break;
+ case 2:
+ printk(KERN_NULL "<%d>%s0x%x\n", gts->printk_level,
+ gts->printk_str->src, (int)(*(short *)pbuf));
+ break;
+ case 4:
+ printk(KERN_NULL "<%d>%s0x%x\n", gts->printk_level,
+ gts->printk_str->src, *(int *)pbuf);
+ break;
+ case 8:
+ printk(KERN_NULL "<%d>%s0x%llx\n", gts->printk_level,
+ gts->printk_str->src, *(long long *)pbuf);
+ break;
+ default:
+ printk(KERN_WARNING "gtp_action_printk: id:%d addr:%p "
+ "size %d cannot printk.\n",
+ (int)gts->tpe->num,
+ (void *)(CORE_ADDR)gts->tpe->addr,
+ (unsigned int)size);
+ gts->tpe->reason = gtp_stop_agent_expr_code_error;
+ return -1;
+ break;
+ }
+ break;
+ case GTP_PRINTK_FORMAT_S:
+ pbuf[GTP_PRINTF_MAX - 1] = '\0';
+ printk("<%d>%s%s\n", gts->printk_level, gts->printk_str->src,
+ pbuf);
+ break;
+ case GTP_PRINTK_FORMAT_B: {
+ size_t i;
+
+ printk(KERN_NULL "<%d>%s", gts->printk_level,
+ gts->printk_str->src);
+ for (i = 0; i < size; i++)
+ printk("%02x", (unsigned int)pbuf[i]);
+ printk("\n");
+ }
+ break;
+ default:
+ printk(KERN_WARNING "gtp_action_printk: id:%d addr:%p "
+ "printk format %u is not support.\n",
+ (int)gts->tpe->num, (void *)(CORE_ADDR)gts->tpe->addr,
+ gts->printk_format);
+ gts->tpe->reason = gtp_stop_agent_expr_code_error;
+ return -1;
+ break;
+ }
+
+ gts->printk_str = gts->printk_str->next;
+
+ return 0;
+}
+
+static int
+gtp_action_memory_read(struct gtp_trace_s *gts, int reg, CORE_ADDR addr,
+ size_t size)
+{
+ char *tmp;
+ struct gtp_frame_mem *fm;
+#ifndef GTP_FRAME_SIMPLE
+ struct ring_buffer_event *rbe;
+#endif
+
+ if (reg >= 0)
+ addr += (CORE_ADDR) gtp_action_reg_read(gts->regs,
+ gts->tpe, reg);
+ if (gts->tpe->reason != gtp_stop_normal)
+ return -1;
+
+ if (gts->next == NULL) {
+ if (gtp_action_head(gts))
+ return -1;
+ }
+
+#ifdef GTP_FRAME_SIMPLE
+ tmp = gtp_frame_alloc(GTP_FRAME_MEM_SIZE + size);
+ if (!tmp) {
+ gts->tpe->reason = gtp_stop_frame_full;
+ return -1;
+ }
+ *gts->next = tmp;
+#else
+ GTP_FRAME_RINGBUFFER_ALLOC(GTP_FRAME_MEM_SIZE + size);
+#endif
+
+ FID(tmp) = FID_MEM;
+ tmp += FID_SIZE;
+
+#ifdef GTP_FRAME_SIMPLE
+ gts->next = (char **)tmp;
+ *gts->next = NULL;
+ tmp += sizeof(char *);
+#endif
+
+ fm = (struct gtp_frame_mem *)tmp;
+ fm->addr = addr;
+ fm->size = size;
+ tmp += sizeof(struct gtp_frame_mem);
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_action_memory_read: id:%d addr:%p %p %u\n",
+ (int)gts->tpe->num, (void *)(CORE_ADDR)gts->tpe->addr,
+ (void *)addr, (unsigned int)size);
+#endif
+
+ if (probe_kernel_read(tmp, (void *)addr, size)) {
+ gts->tpe->reason = gtp_stop_efault;
+#ifdef GTP_FRAME_SIMPLE
+ memset(tmp, 0, size);
+#else
+ ring_buffer_discard_commit(gtp_frame, rbe);
+#endif
+ printk(KERN_WARNING "gtp_action_memory_read: id:%d addr:%p "
+ "read %p %u get error.\n",
+ (int)gts->tpe->num, (void *)(CORE_ADDR)gts->tpe->addr,
+ (void *)addr, (unsigned int)size);
+ return -1;
+ }
+
+#ifndef GTP_FRAME_SIMPLE
+ ring_buffer_unlock_commit(gtp_frame, rbe);
+#endif
+
+ return 0;
+}
+
+static int
+gtp_action_r(struct gtp_trace_s *gts, struct action *ae)
+{
+ struct pt_regs *regs;
+ char *tmp;
+#ifndef GTP_FRAME_SIMPLE
+ struct ring_buffer_event *rbe;
+#endif
+
+ if (gts->next == NULL) {
+ if (gtp_action_head(gts))
+ return -1;
+ }
+
+#ifdef GTP_FRAME_SIMPLE
+ tmp = gtp_frame_alloc(GTP_FRAME_REG_SIZE);
+ if (!tmp) {
+ gts->tpe->reason = gtp_stop_frame_full;
+ return -1;
+ }
+ *gts->next = tmp;
+#else
+ GTP_FRAME_RINGBUFFER_ALLOC(GTP_FRAME_REG_SIZE);
+#endif
+
+ FID(tmp) = FID_REG;
+ tmp += FID_SIZE;
+
+#ifdef GTP_FRAME_SIMPLE
+ gts->next = (char **)tmp;
+ *gts->next = NULL;
+ tmp += sizeof(char *);
+#endif
+
+ regs = (struct pt_regs *)tmp;
+
+ memcpy(regs, gts->regs, sizeof(struct pt_regs));
+#ifdef CONFIG_X86_32
+ regs->sp = (unsigned long)®s->sp;
+#endif /* CONFIG_X86_32 */
+
+ if (gts->ri)
+ GTP_REGS_PC(regs) = (CORE_ADDR)gts->ri->ret_addr;
+#ifdef CONFIG_X86
+ else
+ GTP_REGS_PC(regs) -= 1;
+#endif /* CONFIG_X86 */
+
+#ifndef GTP_FRAME_SIMPLE
+ ring_buffer_unlock_commit(gtp_frame, rbe);
+#endif
+
+ return 0;
+}
+
+static struct gtp_var *
+gtp_gtp_var_array_find(unsigned int num)
+{
+ struct gtp_var *ret;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gtp_var_array_find: num:%u %u %u\n",
+ gtp_var_head, gtp_var_tail, num);
+#endif
+
+ if (num < gtp_var_head || num > gtp_var_tail)
+ return NULL;
+
+ ret = gtp_var_array[num - gtp_var_head];
+ if (ret->per_cpu)
+ ret = ret->per_cpu[smp_processor_id()];
+
+ return ret;
+}
+
+uint64_t
+gtp_get_var(struct gtp_trace_s *gts, struct gtp_var *tve)
+{
+ switch (tve->num) {
+ case GTP_VAR_CURRENT_TASK_ID:
+ if (gts->ri)
+ return (uint64_t)(CORE_ADDR)gts->ri->task;
+ else
+ return (uint64_t)(CORE_ADDR)get_current();
+ break;
+ case GTP_VAR_CURRENT_THREAD_INFO_ID:
+ return (uint64_t)(CORE_ADDR)current_thread_info();
+ break;
+ case GTP_VAR_CLOCK_ID:
+ return (uint64_t)GTP_LOCAL_CLOCK;
+ break;
+ case GTP_VAR_COOKED_CLOCK_ID:
+ return (uint64_t)(__get_cpu_var(local_clock_current)
+ - __get_cpu_var(local_clock_offset));
+ break;
+ case GTP_VAR_COOKED_CLOCK_ID:
+ return (uint64_t)(__get_cpu_var(local_clock_current)
+ - __get_cpu_var(local_clock_offset));
+ break;
+#ifdef CONFIG_X86
+ case GTP_VAR_RDTSC_ID:
+ {
+ unsigned long long a;
+ rdtscll(a);
+ return (uint64_t)a;
+ }
+ break;
+ case GTP_VAR_COOKED_RDTSC_ID:
+ return (uint64_t)(__get_cpu_var(rdtsc_current)
+ - __get_cpu_var(rdtsc_offset));
+ break;
+ case GTP_VAR_COOKED_RDTSC_ID:
+ return (uint64_t)(__get_cpu_var(rdtsc_current)
+ - __get_cpu_var(rdtsc_offset));
+ break;
+#endif
+ case GTP_VAR_CPU_ID:
+ return (uint64_t)(CORE_ADDR)smp_processor_id();
+ break;
+ case GTP_VAR_CPU_NUMBER_ID:
+ return (uint64_t)gtp_cpu_number;
+ break;
+ case GTP_VAR_PRINTK_TMP_ID:
+ return gts->printk_tmp;
+ break;
+ case GTP_VAR_DUMP_STACK_ID:
+ printk(KERN_NULL "gtp %d %p:", (int)gts->tpe->num,
+ (void *)(CORE_ADDR)gts->tpe->addr);
+ dump_stack();
+ return 0;
+ break;
+ case GTP_VAR_XTIME_SEC_ID:
+ if (gts->xtime.tv_sec == 0 && gts->xtime.tv_nsec == 0)
+ getnstimeofday(>s->xtime);
+ return (uint64_t)gts->xtime.tv_sec;
+ break;
+ case GTP_VAR_XTIME_NSEC_ID:
+ if (gts->xtime.tv_sec == 0 && gts->xtime.tv_nsec == 0)
+ getnstimeofday(>s->xtime);
+ return (uint64_t)gts->xtime.tv_nsec;
+ break;
+ case GTP_VAR_HARDIRQ_COUNT_ID:
+ return (uint64_t)hardirq_count();
+ break;
+ case GTP_VAR_SOFTIRQ_COUNT_ID:
+ return (uint64_t)softirq_count();
+ break;
+ case GTP_VAR_IRQ_COUNT_ID:
+ return (uint64_t)irq_count();
+ break;
+ }
+
+#ifdef CONFIG_PERF_EVENTS
+ if (tve->ptid == pe_tv_val || tve->ptid == pe_tv_enabled
+ || tve->ptid == pe_tv_running) {
+ tve->pts->val = perf_event_read_value(tve->pts->event,
+ &(tve->pts->enabled),
+ &(tve->pts->running));
+ switch (tve->ptid) {
+ case pe_tv_val:
+ return (uint64_t)(tve->pts->val);
+ break;
+ case pe_tv_enabled:
+ return (uint64_t)(tve->pts->enabled);
+ break;
+ case pe_tv_running:
+ return (uint64_t)(tve->pts->running);
+ break;
+ default:
+ return 0;
+ break;
+ }
+ }
+#endif
+
+ return tve->val;
+}
+
+static int
+gtp_collect_var(struct gtp_trace_s *gts, struct gtp_var *tve)
+{
+ struct gtp_frame_var *fvar;
+ char *tmp;
+#ifndef GTP_FRAME_SIMPLE
+ struct ring_buffer_event *rbe;
+#endif
+
+ if (gts->next == NULL) {
+ if (gtp_action_head(gts))
+ return -1;
+ }
+
+#ifdef GTP_FRAME_SIMPLE
+ tmp = gtp_frame_alloc(GTP_FRAME_VAR_SIZE);
+ if (!tmp) {
+ gts->tpe->reason = gtp_stop_frame_full;
+ return -1;
+ }
+ *gts->next = tmp;
+#else
+ GTP_FRAME_RINGBUFFER_ALLOC(GTP_FRAME_VAR_SIZE);
+#endif
+
+ FID(tmp) = FID_VAR;
+ tmp += FID_SIZE;
+
+#ifdef GTP_FRAME_SIMPLE
+ gts->next = (char **)tmp;
+ *gts->next = NULL;
+ tmp += sizeof(char *);
+#endif
+
+ fvar = (struct gtp_frame_var *) tmp;
+ fvar->num = tve->num;
+ fvar->val = gtp_get_var(gts, tve);
+
+#ifndef GTP_FRAME_SIMPLE
+ ring_buffer_unlock_commit(gtp_frame, rbe);
+#endif
+
+ return 0;
+}
+
+#define gtp_action_x_getv \
+ do { \
+ struct gtp_var *tve; \
+ \
+ tve = gtp_gtp_var_array_find(arg); \
+ if (!tve) \
+ goto code_error_out; \
+ \
+ stack[sp++] = top; \
+ \
+ top = gtp_get_var(gts, tve); \
+ } while (0)
+
+#ifdef CONFIG_PERF_EVENTS
+#define gtp_action_x_setv_pe \
+ do { \
+ if (tve->ptid == pe_tv_en) \
+ gtp_pe_set_en(tve->pts, (int)top); \
+ else if (tve->ptid == pe_tv_val) \
+ perf_event_set(tve->pts->event, (u64)top); \
+ } while (0)
+#else
+#define gtp_action_x_setv_pe
+#endif
+
+#define gtp_action_x_setv \
+ do { \
+ switch (arg) { \
+ case GTP_VAR_PRINTK_TMP_ID: \
+ gts->printk_tmp = top; \
+ break; \
+ case GTP_VAR_PRINTK_LEVEL_ID: \
+ gts->printk_level = (unsigned int)top; \
+ break; \
+ case GTP_VAR_PRINTK_FORMAT_ID: \
+ gts->printk_format = (unsigned int)top; \
+ break; \
+ case GTP_VAR_PC_PE_EN_ID: \
+ gtp_pc_pe_en((int)top); \
+ break; \
+ default: { \
+ struct gtp_var *tve; \
+ \
+ tve = gtp_gtp_var_array_find(arg); \
+ if (!tve) \
+ goto code_error_out; \
+ gtp_action_x_setv_pe; \
+ /* Not check the other special \
+ trace state variables. \
+ Checked in gtp_check_x. */ \
+ tve->val = (uint64_t)top; \
+ } \
+ break; \
+ } \
+ } while (0)
+
+#define gtp_action_x_tracev \
+ do { \
+ if (!gts->tpe->have_printk \
+ || !GTP_VAR_AUTO_TRACEV(arg)) { \
+ struct gtp_var *tve; \
+ \
+ tve = gtp_gtp_var_array_find(arg); \
+ if (!tve) \
+ goto code_error_out; \
+ \
+ if (gtp_collect_var(gts, tve)) { \
+ /* gtp_collect_var will set error \
+ status with itself if it got \
+ error. */ \
+ goto out; \
+ } \
+ } \
+ } while (0)
+
+#define gtp_action_x_tracev_printk \
+ do { \
+ struct gtp_var *tve; \
+ \
+ tve = gtp_gtp_var_array_find(arg); \
+ if (!tve) \
+ goto code_error_out; \
+ \
+ if (gtp_action_printk(gts, gtp_get_var(gts, tve), 0)) { \
+ /* gtp_collect_var will set error status with \
+ itself if it got error. */ \
+ goto out; \
+ } \
+ } while (0)
+
+#define gtp_action_x_printf \
+ do { \
+ if (strstr((char *)(ebuf + pc), "%s")) { \
+ int i; \
+ char buf[50]; \
+ \
+ for (i = 0; i < 50; i++) { \
+ if (probe_kernel_read(buf + i, \
+ argv + i, 1)) \
+ goto code_error_out; \
+ if (!buf[i]) \
+ break; \
+ } \
+ snprintf(pbuf, psize, (char *)(ebuf + pc), \
+ buf); \
+ } else { \
+ snprintf(pbuf, psize, (char *)(ebuf + pc), \
+ argv); \
+ } \
+ } while (0)
+
+#define STACK_MAX 32
+static DEFINE_PER_CPU(ULONGEST[STACK_MAX], action_x_stack);
+
+static int
+gtp_action_x(struct gtp_trace_s *gts, struct action *ae)
+{
+ int ret = 0;
+ unsigned int pc = 0, sp = 0;
+ ULONGEST top = 0;
+ int arg;
+ union {
+ union {
+ uint8_t bytes[1];
+ uint8_t val;
+ } u8;
+ union {
+ uint8_t bytes[2];
+ uint16_t val;
+ } u16;
+ union {
+ uint8_t bytes[4];
+ uint32_t val;
+ } u32;
+ union {
+ uint8_t bytes[8];
+ ULONGEST val;
+ } u64;
+ } cnv;
+ uint8_t *ebuf = ae->u.exp.buf;
+ int psize = GTP_PRINTF_MAX;
+ char *pbuf = __get_cpu_var(gtp_printf);
+ ULONGEST *stack = __get_cpu_var(action_x_stack);
+
+ if (ae->u.exp.need_var_lock)
+ spin_lock(>p_var_lock);
+
+ if (ae->type == 'X') {
+ while (pc < ae->u.exp.size) {
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_parse_x: cmd %x\n", ebuf[pc]);
+#endif
+
+ switch (ebuf[pc++]) {
+ /* add */
+ case 0x02:
+ top += stack[--sp];
+ break;
+ /* sub */
+ case 0x03:
+ top = stack[--sp] - top;
+ break;
+ /* mul */
+ case 0x04:
+ top *= stack[--sp];
+ break;
+#ifndef CONFIG_MIPS
+ /* div_signed */
+ case 0x05:
+ if (top) {
+ LONGEST l = (LONGEST) stack[--sp];
+ do_div(l, (LONGEST) top);
+ top = l;
+ } else
+ goto code_error_out;
+ break;
+ /* div_unsigned */
+ case 0x06:
+ if (top) {
+ ULONGEST ul = stack[--sp];
+ do_div(ul, top);
+ top = ul;
+ } else
+ goto code_error_out;
+ break;
+ /* rem_signed */
+ case 0x07:
+ if (top) {
+ LONGEST l1 = (LONGEST) stack[--sp];
+ LONGEST l2 = (LONGEST) top;
+ top = do_div(l1, l2);
+ } else
+ goto code_error_out;
+ break;
+ /* rem_unsigned */
+ case 0x08:
+ if (top) {
+ ULONGEST ul1 = stack[--sp];
+ ULONGEST ul2 = top;
+ top = do_div(ul1, ul2);
+ } else
+ goto code_error_out;
+ break;
+#endif
+ /* lsh */
+ case 0x09:
+ top = stack[--sp] << top;
+ break;
+ /* rsh_signed */
+ case 0x0a:
+ top = ((LONGEST) stack[--sp]) >> top;
+ break;
+ /* rsh_unsigned */
+ case 0x0b:
+ top = stack[--sp] >> top;
+ break;
+ /* trace */
+ case 0x0c:
+ --sp;
+ if (!gts->tpe->have_printk) {
+ if (gtp_action_memory_read
+ (gts, -1,
+ (CORE_ADDR) stack[sp],
+ (size_t) top))
+ goto out;
+ }
+ top = stack[--sp];
+ break;
+ /* trace_printk */
+ case 0xfd:
+ if (gtp_action_printk(gts,
+ (ULONGEST)stack[--sp],
+ (size_t) top))
+ goto out;
+ top = stack[--sp];
+ break;
+ /* trace_quick */
+ case 0x0d:
+ if (!gts->tpe->have_printk) {
+ if (gtp_action_memory_read
+ (gts, -1, (CORE_ADDR) top,
+ (size_t) ebuf[pc]))
+ goto out;
+ }
+ pc++;
+ break;
+ /* trace_quick_printk */
+ case 0xfe:
+ if (gtp_action_printk(gts, (ULONGEST) top,
+ (size_t) ebuf[pc++]))
+ goto out;
+ break;
+ /* log_not */
+ case 0x0e:
+ top = !top;
+ break;
+ /* bit_and */
+ case 0x0f:
+ top &= stack[--sp];
+ break;
+ /* bit_or */
+ case 0x10:
+ top |= stack[--sp];
+ break;
+ /* bit_xor */
+ case 0x11:
+ top ^= stack[--sp];
+ break;
+ /* bit_not */
+ case 0x12:
+ top = ~top;
+ break;
+ /* equal */
+ case 0x13:
+ top = (stack[--sp] == top);
+ break;
+ /* less_signed */
+ case 0x14:
+ top = (((LONGEST) stack[--sp])
+ < ((LONGEST) top));
+ break;
+ /* less_unsigned */
+ case 0x15:
+ top = (stack[--sp] < top);
+ break;
+ /* ext */
+ case 0x16:
+ arg = ebuf[pc++];
+ if (arg < (sizeof(LONGEST)*8)) {
+ LONGEST mask = 1 << (arg - 1);
+ top &= ((LONGEST) 1 << arg) - 1;
+ top = (top ^ mask) - mask;
+ }
+ break;
+ /* ref8 */
+ case 0x17:
+ if (probe_kernel_read
+ (cnv.u8.bytes,
+ (void *)(CORE_ADDR)top, 1))
+ goto code_error_out;
+ top = (ULONGEST) cnv.u8.val;
+ break;
+ /* ref16 */
+ case 0x18:
+ if (probe_kernel_read
+ (cnv.u16.bytes,
+ (void *)(CORE_ADDR)top, 2))
+ goto code_error_out;
+ top = (ULONGEST) cnv.u16.val;
+ break;
+ /* ref32 */
+ case 0x19:
+ if (probe_kernel_read
+ (cnv.u32.bytes,
+ (void *)(CORE_ADDR)top, 4))
+ goto code_error_out;
+ top = (ULONGEST) cnv.u32.val;
+ break;
+ /* ref64 */
+ case 0x1a:
+ if (probe_kernel_read
+ (cnv.u64.bytes,
+ (void *)(CORE_ADDR)top, 8))
+ goto code_error_out;
+ top = (ULONGEST) cnv.u64.val;
+ break;
+ /* if_goto */
+ case 0x20:
+ /* The not check sp code don't
+ support if_goto. */
+ goto code_error_out;
+ break;
+ /* goto */
+ case 0x21:
+ pc = (ebuf[pc] << 8) + (ebuf[pc + 1]);
+ break;
+ /* const8 */
+ case 0x22:
+ stack[sp++] = top;
+ top = ebuf[pc++];
+ break;
+ /* const16 */
+ case 0x23:
+ stack[sp++] = top;
+ top = ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ break;
+ /* const32 */
+ case 0x24:
+ stack[sp++] = top;
+ top = ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ break;
+ /* const64 */
+ case 0x25:
+ stack[sp++] = top;
+ top = ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ break;
+ /* reg */
+ case 0x26:
+ stack[sp++] = top;
+ arg = ebuf[pc++];
+ arg = (arg << 8) + ebuf[pc++];
+ top = gtp_action_reg_read(gts->regs, gts->tpe,
+ arg);
+ if (gts->tpe->reason != gtp_stop_normal)
+ goto error_out;
+ break;
+ /* end */
+ case 0x27:
+ if (gts->run)
+ *(gts->run) = (int)top;
+ goto out;
+ break;
+ /* dup */
+ case 0x28:
+ stack[sp++] = top;
+ break;
+ /* pop */
+ case 0x29:
+ top = stack[--sp];
+ break;
+ /* zero_ext */
+ case 0x2a:
+ arg = ebuf[pc++];
+ if (arg < (sizeof(LONGEST)*8))
+ top &= ((LONGEST) 1 << arg) - 1;
+ break;
+ /* swap */
+ case 0x2b:
+ stack[sp] = top;
+ top = stack[sp - 1];
+ stack[sp - 1] = stack[sp];
+ break;
+ /* getv */
+ case 0x2c:
+ arg = ebuf[pc++];
+ arg = (arg << 8) + ebuf[pc++];
+ gtp_action_x_getv;
+ break;
+ /* setv */
+ case 0x2d:
+ arg = ebuf[pc++];
+ arg = (arg << 8) + ebuf[pc++];
+ gtp_action_x_setv;
+ break;
+ /* tracev */
+ case 0x2e:
+ arg = ebuf[pc++];
+ arg = (arg << 8) + ebuf[pc++];
+ gtp_action_x_tracev;
+ break;
+ /* tracev_printk */
+ case 0xff:
+ arg = ebuf[pc++];
+ arg = (arg << 8) + ebuf[pc++];
+ gtp_action_x_tracev_printk;
+ break;
+ /* printf */
+ case 0x31: {
+ arg = ebuf[pc++];
+
+ if (arg) {
+ void *argv = (void *)
+ (unsigned long)
+ top;
+
+ /* pop */
+ top = stack[--sp];
+
+ gtp_action_x_printf;
+ } else
+ snprintf(pbuf, psize,
+ (char *)(ebuf + pc));
+ psize -= strlen(pbuf);
+ pbuf += strlen(pbuf);
+
+ pc += strlen((char *)ebuf + pc) + 1;
+ }
+ break;
+ }
+ }
+ } else {
+ /* The x execution code don't support printk so it doesn't have
+ printk ae support. */
+ while (pc < ae->u.exp.size) {
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_parse_x: cmd %x\n", ebuf[pc]);
+#endif
+
+ switch (ebuf[pc++]) {
+ /* add */
+ case 0x02:
+ if (sp)
+ top += stack[--sp];
+ else
+ goto code_error_out;
+ break;
+ /* sub */
+ case 0x03:
+ if (sp)
+ top = stack[--sp] - top;
+ else
+ goto code_error_out;
+ break;
+ /* mul */
+ case 0x04:
+ if (sp)
+ top *= stack[--sp];
+ else
+ goto code_error_out;
+ break;
+#ifndef CONFIG_MIPS
+ /* div_signed */
+ case 0x05:
+ if (top && sp) {
+ LONGEST l = (LONGEST) stack[--sp];
+ do_div(l, (LONGEST) top);
+ top = l;
+ } else
+ goto code_error_out;
+ break;
+ /* div_unsigned */
+ case 0x06:
+ if (top && sp) {
+ ULONGEST ul = stack[--sp];
+ do_div(ul, top);
+ top = ul;
+ } else
+ goto code_error_out;
+ break;
+ /* rem_signed */
+ case 0x07:
+ if (top && sp) {
+ LONGEST l1 = (LONGEST) stack[--sp];
+ LONGEST l2 = (LONGEST) top;
+ top = do_div(l1, l2);
+ } else
+ goto code_error_out;
+ break;
+ /* rem_unsigned */
+ case 0x08:
+ if (top && sp) {
+ ULONGEST ul1 = stack[--sp];
+ ULONGEST ul2 = top;
+ top = do_div(ul1, ul2);
+ } else
+ goto code_error_out;
+ break;
+#endif
+ /* lsh */
+ case 0x09:
+ if (sp)
+ top = stack[--sp] << top;
+ else
+ goto code_error_out;
+ break;
+ /* rsh_signed */
+ case 0x0a:
+ if (sp)
+ top = ((LONGEST) stack[--sp]) >> top;
+ else
+ goto code_error_out;
+ break;
+ /* rsh_unsigned */
+ case 0x0b:
+ if (sp)
+ top = stack[--sp] >> top;
+ else
+ goto code_error_out;
+ break;
+ /* trace */
+ case 0x0c:
+ if (sp > 1) {
+ if (gtp_action_memory_read
+ (gts, -1, (CORE_ADDR) stack[--sp],
+ (size_t) top)) {
+ /* gtp_action_memory_read will
+ set error status with itself
+ if it got error. */
+ goto out;
+ }
+ top = stack[--sp];
+ } else
+ goto code_error_out;
+ break;
+ /* trace_quick */
+ case 0x0d:
+ if (gtp_action_memory_read
+ (gts, -1, (CORE_ADDR) top,
+ (size_t) ebuf[pc++])) {
+ /* gtp_action_memory_read will set
+ error status with itself if it got
+ error. */
+ goto out;
+ }
+ break;
+ /* log_not */
+ case 0x0e:
+ top = !top;
+ break;
+ /* bit_and */
+ case 0x0f:
+ if (sp)
+ top &= stack[--sp];
+ else
+ goto code_error_out;
+ break;
+ /* bit_or */
+ case 0x10:
+ if (sp)
+ top |= stack[--sp];
+ else
+ goto code_error_out;
+ break;
+ /* bit_xor */
+ case 0x11:
+ if (sp)
+ top ^= stack[--sp];
+ else
+ goto code_error_out;
+ break;
+ /* bit_not */
+ case 0x12:
+ top = ~top;
+ break;
+ /* equal */
+ case 0x13:
+ if (sp)
+ top = (stack[--sp] == top);
+ else
+ goto code_error_out;
+ break;
+ /* less_signed */
+ case 0x14:
+ if (sp)
+ top = (((LONGEST) stack[--sp])
+ < ((LONGEST) top));
+ else
+ goto code_error_out;
+ break;
+ /* less_unsigned */
+ case 0x15:
+ if (sp)
+ top = (stack[--sp] < top);
+ else
+ goto code_error_out;
+ break;
+ /* ext */
+ case 0x16:
+ arg = ebuf[pc++];
+ if (arg < (sizeof(LONGEST)*8)) {
+ LONGEST mask = 1 << (arg - 1);
+ top &= ((LONGEST) 1 << arg) - 1;
+ top = (top ^ mask) - mask;
+ }
+ break;
+ /* ref8 */
+ case 0x17:
+ if (probe_kernel_read
+ (cnv.u8.bytes,
+ (void *)(CORE_ADDR)top, 1))
+ goto code_error_out;
+ top = (ULONGEST) cnv.u8.val;
+ break;
+ /* ref16 */
+ case 0x18:
+ if (probe_kernel_read
+ (cnv.u16.bytes,
+ (void *)(CORE_ADDR)top, 2))
+ goto code_error_out;
+ top = (ULONGEST) cnv.u16.val;
+ break;
+ /* ref32 */
+ case 0x19:
+ if (probe_kernel_read
+ (cnv.u32.bytes,
+ (void *)(CORE_ADDR)top, 4))
+ goto code_error_out;
+ top = (ULONGEST) cnv.u32.val;
+ break;
+ /* ref64 */
+ case 0x1a:
+ if (probe_kernel_read
+ (cnv.u64.bytes,
+ (void *)(CORE_ADDR)top, 8))
+ goto code_error_out;
+ top = (ULONGEST) cnv.u64.val;
+ break;
+ /* if_goto */
+ case 0x20:
+ if (top)
+ pc = (ebuf[pc] << 8)
+ + (ebuf[pc + 1]);
+ else
+ pc += 2;
+ /* pop */
+ if (sp)
+ top = stack[--sp];
+ else
+ goto code_error_out;
+ break;
+ /* goto */
+ case 0x21:
+ pc = (ebuf[pc] << 8) + (ebuf[pc + 1]);
+ break;
+ /* const8 */
+ case 0x22:
+ stack[sp++] = top;
+ top = ebuf[pc++];
+ break;
+ /* const16 */
+ case 0x23:
+ stack[sp++] = top;
+ top = ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ break;
+ /* const32 */
+ case 0x24:
+ stack[sp++] = top;
+ top = ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ break;
+ /* const64 */
+ case 0x25:
+ stack[sp++] = top;
+ top = ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ top = (top << 8) + ebuf[pc++];
+ break;
+ /* reg */
+ case 0x26:
+ stack[sp++] = top;
+ arg = ebuf[pc++];
+ arg = (arg << 8) + ebuf[pc++];
+ top = gtp_action_reg_read(gts->regs, gts->tpe,
+ arg);
+ if (gts->tpe->reason != gtp_stop_normal)
+ goto error_out;
+ break;
+ /* end */
+ case 0x27:
+ if (gts->run)
+ *(gts->run) = (int)top;
+ goto out;
+ break;
+ /* dup */
+ case 0x28:
+ stack[sp++] = top;
+ break;
+ /* pop */
+ case 0x29:
+ if (sp)
+ top = stack[--sp];
+ else
+ goto code_error_out;
+ break;
+ /* zero_ext */
+ case 0x2a:
+ arg = ebuf[pc++];
+ if (arg < (sizeof(LONGEST)*8))
+ top &= ((LONGEST) 1 << arg) - 1;
+ break;
+ /* swap */
+ case 0x2b:
+ if (sp) {
+ stack[sp] = top;
+ top = stack[sp - 1];
+ stack[sp - 1] = stack[sp];
+ } else
+ goto code_error_out;
+ break;
+ /* getv */
+ case 0x2c:
+ arg = ebuf[pc++];
+ arg = (arg << 8) + ebuf[pc++];
+ if (GTP_VAR_NOT_GETV(arg))
+ goto code_error_out;
+ gtp_action_x_getv;
+ break;
+ /* setv */
+ case 0x2d:
+ arg = ebuf[pc++];
+ arg = (arg << 8) + ebuf[pc++];
+ if (GTP_VAR_NOT_SETV(arg))
+ goto code_error_out;
+ gtp_action_x_setv;
+ break;
+ /* tracev */
+ case 0x2e:
+ arg = ebuf[pc++];
+ arg = (arg << 8) + ebuf[pc++];
+ if (GTP_VAR_NOT_TRACEV(arg))
+ goto code_error_out;
+ gtp_action_x_tracev;
+ break;
+ /* printf */
+ case 0x31: {
+ arg = ebuf[pc++];
+
+ if (arg) {
+ void *argv = (void *)
+ (unsigned long)
+ top;
+
+ /* pop */
+ if (sp)
+ top = stack[--sp];
+ else
+ goto code_error_out;
+
+ gtp_action_x_printf;
+ } else
+ snprintf(pbuf, psize,
+ (char *)(ebuf + pc));
+ psize -= strlen(pbuf);
+ pbuf += strlen(pbuf);
+
+ pc += strlen((char *)ebuf + pc) + 1;
+ }
+ break;
+ }
+
+ if (sp > STACK_MAX - 5) {
+ printk(KERN_WARNING "gtp_action_x: stack "
+ "overflow.\n");
+ gts->tpe->reason
+ = gtp_stop_agent_expr_stack_overflow;
+ goto error_out;
+ }
+ }
+ }
+code_error_out:
+ gts->tpe->reason = gtp_stop_agent_expr_code_error;
+error_out:
+ ret = -1;
+ printk(KERN_WARNING "gtp_action_x: tracepoint %d addr:%p"
+ "action X get error in pc %u.\n",
+ (int)gts->tpe->num, (void *)(CORE_ADDR)gts->tpe->addr, pc);
+out:
+ if (psize != GTP_PRINTF_MAX) {
+ unsigned long flags;
+
+ local_irq_save(flags);
+ printk("%s", pbuf - (GTP_PRINTF_MAX - psize));
+ local_irq_restore(flags);
+ }
+ if (ae->u.exp.need_var_lock)
+ spin_unlock(>p_var_lock);
+ return ret;
+}
+
+#ifndef GTP_FRAME_SIMPLE
+static void
+gtp_kp_pre_handler_end(void)
+{
+ FID_TYPE eid = FID_END;
+ ring_buffer_write(gtp_frame, FID_SIZE, &eid);
+
+ if (atomic_read(>pframe_pipe_wq_v) > 0) {
+ atomic_dec(>pframe_pipe_wq_v);
+ add_preempt_count(HARDIRQ_OFFSET);
+ tasklet_schedule(>pframe_pipe_wq_tasklet);
+ sub_preempt_count(HARDIRQ_OFFSET);
+ }
+}
+#endif
+
+static int
+gtp_handler(struct gtp_trace_s *gts)
+{
+ struct action *ae;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_handler: tracepoint %d %p\n",
+ (int)gts->tpe->num, (void *)(CORE_ADDR)gts->tpe->addr);
+#endif
+
+ if (gts->tpe->kpreg == 0)
+ return 0;
+
+ if (gts->tpe->no_self_trace) {
+ if (get_current()->pid == gtp_gtp_pid
+ || get_current()->pid == gtp_gtpframe_pid)
+ return 0;
+#ifndef GTP_FRAME_SIMPLE
+ if (get_current()->pid == gtp_gtpframe_pipe_pid)
+ return 0;
+#endif
+ }
+
+ if (gts->tpe->have_printk) {
+ gts->printk_level = 8;
+ gts->printk_str = gts->tpe->printk_str;
+ }
+
+ /* Condition. */
+ if (gts->tpe->cond) {
+ int run;
+
+ gts->run = &run;
+ if (gtp_action_x(gts, gts->tpe->cond))
+ goto tpe_stop;
+ if (!run)
+ return 0;
+ }
+
+ gts->run = NULL;
+
+ /* Pass. */
+ if (!gts->tpe->nopass) {
+ if (atomic_dec_return(>s->tpe->current_pass) < 0)
+ goto tpe_stop;
+ }
+
+ atomic_inc(>p_frame_create);
+
+ /* Handle actions. */
+ for (ae = gts->tpe->action_list; ae; ae = ae->next) {
+ switch (ae->type) {
+ case 'R':
+ if (gtp_action_r(gts, ae))
+ goto tpe_stop;
+ break;
+ case 'X':
+ case 0xff:
+ if (gtp_action_x(gts, ae))
+ goto tpe_stop;
+ break;
+ case 'M':
+ if (gtp_action_memory_read(gts, ae->u.m.regnum,
+ ae->u.m.offset,
+ ae->u.m.size))
+ goto tpe_stop;
+ break;
+ }
+ }
+
+#ifndef GTP_FRAME_SIMPLE
+ if (gts->next)
+ gtp_kp_pre_handler_end();
+#endif
+
+ return 0;
+
+tpe_stop:
+#ifndef GTP_FRAME_SIMPLE
+ if (gts->next)
+ gtp_kp_pre_handler_end();
+#endif
+ gts->tpe->kpreg = 0;
+ add_preempt_count(HARDIRQ_OFFSET);
+ tasklet_schedule(>s->tpe->tasklet);
+ sub_preempt_count(HARDIRQ_OFFSET);
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_kp_pre_handler: tracepoint %d %p stop.\n",
+ (int)gts->tpe->num, (void *)(CORE_ADDR)gts->tpe->addr);
+#endif
+ return 0;
+}
+
+static void
+gtp_handler_begin(void)
+{
+ if (__get_cpu_var(rdtsc_current) == 0) {
+ u64 a;
+
+ rdtscll(a);
+ __get_cpu_var(rdtsc_current) = a;
+
+ __get_cpu_var(local_clock_current) = GTP_LOCAL_CLOCK;
+#ifdef CONFIG_PERF_EVENTS
+ pc_pe_list_disable();
+#endif
+ }
+}
+
+static void
+gtp_handler_end(void)
+{
+ if (__get_cpu_var(rdtsc_current) != 0) {
+ u64 a;
+
+#ifdef CONFIG_PERF_EVENTS
+ pc_pe_list_enable();
+#endif
+
+ __get_cpu_var(local_clock_offset) += GTP_LOCAL_CLOCK
+ - __get_cpu_var(local_clock_current);
+ __get_cpu_var(local_clock_current) = 0;
+
+ rdtscll(a);
+ __get_cpu_var(rdtsc_offset) += a - __get_cpu_var(rdtsc_current);
+ __get_cpu_var(rdtsc_current) = 0;
+ }
+}
+
+static int
+gtp_kp_pre_handler(struct kprobe *p, struct pt_regs *regs)
+{
+ struct kretprobe *kp;
+ struct gtp_trace_s gts;
+
+ gtp_handler_begin();
+
+ memset(>s, 0, sizeof(struct gtp_trace_s));
+ kp = container_of(p, struct kretprobe, kp);
+ gts.tpe = container_of(kp, struct gtp_entry, kp);
+ gts.regs = regs;
+
+ return gtp_handler(>s);
+}
+
+static void
+gtp_kp_post_handler(struct kprobe *p, struct pt_regs *regs, unsigned
long flags)
+{
+ gtp_handler_end();
+}
+
+static int
+gtp_kp_ret_handler(struct kretprobe_instance *ri, struct pt_regs *regs)
+{
+ int ret;
+ struct gtp_trace_s gts;
+
+ gtp_handler_begin();
+
+ memset(>s, 0, sizeof(struct gtp_trace_s));
+ gts.tpe = container_of(ri->rp, struct gtp_entry, kp);
+ gts.regs = regs;
+ gts.ri = ri;
+
+ ret = gtp_handler(>s);
+
+ gtp_handler_end();
+
+ return ret;
+}
+
+static struct action *
+gtp_action_alloc(char *pkg)
+{
+ struct action *ret;
+
+ ret = kmalloc(sizeof(struct action), GFP_KERNEL);
+ if (!ret)
+ goto out;
+
+ memset(ret, '\0', sizeof(struct action));
+ ret->type = pkg[0];
+ ret->src = pkg;
+
+out:
+ return ret;
+}
+
+static void
+gtp_action_release(struct action *ae)
+{
+ struct action *ae2;
+
+ while (ae) {
+ ae2 = ae;
+ ae = ae->next;
+ /* Release ae2. */
+ switch (ae2->type) {
+ case 'X':
+ case 0xff:
+ kfree(ae2->u.exp.buf);
+ break;
+ }
+ kfree(ae2->src);
+ kfree(ae2);
+ }
+}
+
+static void
+gtp_src_release(struct gtpsrc *src)
+{
+ struct gtpsrc *src2;
+
+ while (src) {
+ src2 = src;
+ src = src->next;
+ kfree(src2->src);
+ kfree(src2);
+ }
+}
+
+static void
+gtp_stop(struct work_struct *work)
+{
+ struct gtp_entry *tpe = container_of(work,
+ struct gtp_entry, work);
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_stop: tracepoint %d %p\n", (int)tpe->num,
+ (void *)(CORE_ADDR)tpe->addr);
+#endif
+
+ if (tpe->is_kretprobe)
+ unregister_kretprobe(&tpe->kp);
+ else
+ unregister_kprobe(&tpe->kp.kp);
+}
+
+static struct gtp_entry *
+gtp_list_add(ULONGEST num, ULONGEST addr)
+{
+ struct gtp_entry *ret = kcalloc(1, sizeof(struct gtp_entry),
+ GFP_KERNEL);
+
+ if (!ret)
+ goto out;
+ memset(ret, '\0', sizeof(struct gtp_entry));
+ ret->num = num;
+ ret->addr = addr;
+ ret->kp.kp.addr = (kprobe_opcode_t *) (CORE_ADDR)addr;
+ INIT_WORK(&ret->work, gtp_stop);
+ ret->have_printk = 0;
+
+ /* Add to gtp_list. */
+ ret->next = gtp_list;
+ gtp_list = ret;
+
+out:
+ return ret;
+}
+
+static struct gtp_entry *
+gtp_list_find(ULONGEST num, ULONGEST addr)
+{
+ struct gtp_entry *tpe;
+
+ for (tpe = gtp_list; tpe; tpe = tpe->next) {
+ if (tpe->num == num && tpe->addr == addr)
+ return tpe;
+ }
+
+ return NULL;
+}
+
+/* If more than one gtp entry have same num, return NULL. */
+
+static struct gtp_entry *
+gtp_list_find_without_addr(ULONGEST num)
+{
+ struct gtp_entry *tpe, *ret = NULL;
+
+ for (tpe = gtp_list; tpe; tpe = tpe->next) {
+ if (tpe->num == num) {
+ if (ret)
+ return NULL;
+ else
+ ret = tpe;
+ }
+ }
+
+ return ret;
+}
+
+static void
+gtp_list_release(void)
+{
+ struct gtp_entry *tpe;
+
+ while (gtp_list) {
+ tpe = gtp_list;
+ gtp_list = gtp_list->next;
+ gtp_action_release(tpe->cond);
+ gtp_action_release(tpe->action_list);
+ gtp_src_release(tpe->src);
+ kfree(tpe);
+ }
+
+ current_gtp = NULL;
+ current_gtp_action = NULL;
+ current_gtp_src = NULL;
+}
+
+#ifndef GTP_FRAME_SIMPLE
+static void
+gtp_frame_iter_open(void)
+{
+ int cpu;
+
+
+ for_each_online_cpu(cpu)
+ gtp_frame_iter[cpu] = ring_buffer_read_prepare(gtp_frame, cpu);
+ ring_buffer_read_prepare_sync();
+ for_each_online_cpu(cpu) {
+ ring_buffer_read_start(gtp_frame_iter[cpu]);
+ }
+}
+
+static void
+gtp_frame_iter_reset(void)
+{
+ int cpu;
+
+ for_each_online_cpu(cpu)
+ ring_buffer_iter_reset(gtp_frame_iter[cpu]);
+ gtp_frame_current_num = -1;
+}
+
+static int
+gtp_frame_iter_peek_head(void)
+{
+ int cpu;
+ int ret = -1;
+ u64 min = 0;
+
+ for_each_online_cpu(cpu) {
+ struct ring_buffer_event *rbe;
+ char *tmp;
+ u64 ts;
+
+ while (1) {
+ rbe = ring_buffer_iter_peek(gtp_frame_iter[cpu], &ts);
+ if (rbe == NULL)
+ break;
+ tmp = ring_buffer_event_data(rbe);
+ if (FID(tmp) == FID_HEAD)
+ break;
+ ring_buffer_read(gtp_frame_iter[cpu], NULL);
+ }
+
+ if (rbe) {
+ if ((min && ts < min) || !min) {
+ min = ts;
+ ret = cpu;
+ }
+ }
+ }
+
+ if (ret < 0)
+ gtp_frame_current_num = -1;
+ else
+ gtp_frame_current_num++;
+ return ret;
+}
+
+static void
+gtp_frame_iter_close(void)
+{
+ int cpu;
+
+ for_each_online_cpu(cpu) {
+ if (gtp_frame_iter[cpu]) {
+ ring_buffer_read_finish(gtp_frame_iter[cpu]);
+ gtp_frame_iter[cpu] = NULL;
+ }
+ }
+}
+#endif
+
+static void
+gtp_frame_reset(void)
+{
+ gtp_frame_current_num = -1;
+#ifdef GTP_FRAME_SIMPLE
+ gtp_frame_r_start = gtp_frame;
+ gtp_frame_w_start = gtp_frame;
+ gtp_frame_end = gtp_frame + GTP_FRAME_SIZE;
+ gtp_frame_is_circular = 0;
+ gtp_frame_current = NULL;
+#else
+ gtp_frame_iter_close();
+ if (gtp_frame)
+ ring_buffer_reset(gtp_frame);
+#endif
+ atomic_set(>p_frame_create, 0);
+ if (gtp_frame_file) {
+ vfree(gtp_frame_file);
+ gtp_frame_file = NULL;
+ gtp_frame_file_size = 0;
+ }
+}
+
+static int
+hex2int(char hex, int *i)
+{
+ if ((hex >= '0') && (hex <= '9')) {
+ *i = hex - '0';
+ return 1;
+ }
+ if ((hex >= 'a') && (hex <= 'f')) {
+ *i = hex - 'a' + 10;
+ return 1;
+ }
+ if ((hex >= 'A') && (hex <= 'F')) {
+ *i = hex - 'A' + 10;
+ return 1;
+ }
+
+ return 0;
+}
+
+static char *
+hex2ulongest(char *pkg, ULONGEST *u64)
+{
+ int i;
+
+ if (u64)
+ *u64 = 0;
+ while (hex2int(pkg[0], &i)) {
+ pkg++;
+ if (u64) {
+ *u64 = (*u64) << 4;
+ *u64 |= i & 0xf;
+ }
+ }
+
+ return pkg;
+}
+
+static char *
+string2hex(char *pkg, char *out)
+{
+ char *ret = out;
+
+ while (pkg[0]) {
+ sprintf(out, "%x", pkg[0]);
+ pkg++;
+ out += 2;
+ }
+
+ return ret;
+}
+
+static char *
+hex2string(char *pkg, char *out)
+{
+ char *ret = out;
+ int i, j;
+
+ while (hex2int(pkg[0], &i) && hex2int(pkg[1], &j)) {
+ out[0] = i * 16 + j;
+ pkg += 2;
+ out += 1;
+ }
+ out[0] = '\0';
+
+ return ret;
+}
+
+static char *
+gtp_strdup(char *begin, char *end)
+{
+ int len;
+ char *ret;
+
+ if (end)
+ len = end - begin;
+ else
+ len = strlen(begin);
+
+ ret = kmalloc(len + 1, GFP_KERNEL);
+ if (ret == NULL)
+ return NULL;
+
+ strncpy(ret, begin, len);
+ ret[len] = '\0';
+
+ return ret;
+}
+
+static void
+gtpro_list_clear(void)
+{
+ struct gtpro_entry *e;
+
+ while (gtpro_list) {
+ e = gtpro_list;
+ gtpro_list = gtpro_list->next;
+ kfree(e);
+ }
+}
+
+static struct gtpro_entry *
+gtpro_list_add(CORE_ADDR start, CORE_ADDR end)
+{
+ struct gtpro_entry *e;
+
+ e = kmalloc(sizeof(struct gtpro_entry), GFP_KERNEL);
+ if (e == NULL)
+ goto out;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtpro_list_add: %p %p\n", (void *)start, (void *)end);
+#endif
+
+ e->start = start;
+ e->end = end;
+
+ e->next = gtpro_list;
+ gtpro_list = e;
+
+out:
+ return e;
+}
+
+#ifdef CONFIG_PERF_EVENTS
+static struct gtp_var *
+gtp_var_add(unsigned int num, uint64_t val, char *src,
+ struct gtp_var **per_cpu, int per_cpu_id,
+ enum pe_tv_id ptid, struct pe_tv_s *pts)
+#else
+static struct gtp_var *
+gtp_var_add(unsigned int num, uint64_t val, char *src,
+ struct gtp_var **per_cpu, int per_cpu_id)
+#endif
+{
+ struct gtp_var *var = kcalloc(1, sizeof(struct gtp_var), GFP_KERNEL);
+ if (!var)
+ goto out;
+
+ var->num = num;
+ var->val = val;
+
+ var->src = gtp_strdup(src, NULL);
+ if (var->src == NULL) {
+ kfree(var);
+ var = NULL;
+ goto out;
+ }
+
+ var->per_cpu = per_cpu;
+ if (per_cpu)
+ var->per_cpu[per_cpu_id] = var;
+
+#ifdef CONFIG_PERF_EVENTS
+ var->ptid = ptid;
+ var->pts = pts;
+#endif
+
+ var->next = gtp_var_list;
+ gtp_var_list = var;
+ gtp_var_head = min(var->num, gtp_var_head);
+ gtp_var_tail = max(var->num, gtp_var_tail);
+
+out:
+ return var;
+}
+
+static struct gtp_var *
+gtp_var_find(unsigned int num)
+{
+ struct gtp_var *ret = NULL;
+
+ if (num >= gtp_var_head && num <= gtp_var_tail) {
+ for (ret = gtp_var_list; ret; ret = ret->next) {
+ if (ret->num == num)
+ break;
+ }
+ }
+
+ return ret;
+}
+
+static void
+gtp_var_release(void)
+{
+ struct gtp_var *tve;
+
+ gtp_var_head = GTP_VAR_SPECIAL_MIN;
+ gtp_var_tail = GTP_VAR_SPECIAL_MAX;
+ current_gtp_var = NULL;
+
+ while (gtp_var_list != GTP_VAR_LIST_FIRST) {
+ tve = gtp_var_list;
+ gtp_var_list = gtp_var_list->next;
+
+ if (tve->per_cpu) {
+ struct gtp_var *tve1;
+
+ for (tve1 = gtp_var_list; tve1; tve1 = tve1->next) {
+ if (tve1->per_cpu == tve->per_cpu)
+ tve1->per_cpu = NULL;
+ }
+
+ kfree(tve->per_cpu);
+ }
+
+#ifdef CONFIG_PERF_EVENTS
+ if (tve->pts) {
+ struct gtp_var *tve1;
+
+ for (tve1 = gtp_var_list; tve1; tve1 = tve1->next) {
+ if (tve1->pts == tve->pts) {
+ tve1->pts = NULL;
+ tve1->ptid = pe_tv_unknown;
+ }
+ }
+
+ if (tve->pts->event)
+ perf_event_release_kernel(tve->pts->event);
+ kfree(tve->pts);
+ }
+#endif
+
+ kfree(tve->src);
+ kfree(tve);
+ }
+
+ gtp_start_ignore_error = 0;
+}
+
+static int
+gtp_gdbrsp_qtstop(void)
+{
+ struct gtp_entry *tpe;
+#ifdef CONFIG_PERF_EVENTS
+ struct gtp_var *tve;
+#endif
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qtstop\n");
+#endif
+
+#ifdef FRAME_ALLOC_RECORD
+ printk(KERN_WARNING "frame_alloc_size = %llu, "
+ "frame_alloc_size_hole = %llu\n",
+ frame_alloc_size, frame_alloc_size_hole);
+ frame_alloc_size = 0;
+ frame_alloc_size_hole = 0;
+#endif
+
+ if (!gtp_start)
+ return -EBUSY;
+
+ flush_workqueue(gtp_wq);
+
+ for (tpe = gtp_list; tpe; tpe = tpe->next) {
+ if (tpe->kpreg) {
+ if (tpe->is_kretprobe)
+ unregister_kretprobe(&tpe->kp);
+ else
+ unregister_kprobe(&tpe->kp.kp);
+ tpe->kpreg = 0;
+ }
+ tasklet_kill(&tpe->tasklet);
+ }
+
+#ifdef CONFIG_PERF_EVENTS
+ for (tve = gtp_var_list; tve; tve = tve->next) {
+ if (tve->pts == NULL)
+ continue;
+ if (tve->pts->event == NULL)
+ continue;
+
+ tve->pts->val = perf_event_read_value(tve->pts->event,
+ &(tve->pts->enabled),
+ &(tve->pts->running));
+ perf_event_release_kernel(tve->pts->event);
+ tve->pts->event = NULL;
+ }
+#endif
+
+ kfree(gtp_var_array);
+ gtp_var_array = NULL;
+
+#ifndef GTP_FRAME_SIMPLE
+ if (gtp_frame) {
+ gtp_frame_iter_open();
+ gtp_frame_iter_reset();
+ }
+#endif
+
+ gtp_start = 0;
+#ifndef GTP_FRAME_SIMPLE
+ if (atomic_read(>pframe_pipe_wq_v) > 0) {
+ atomic_dec(>pframe_pipe_wq_v);
+ tasklet_schedule(>pframe_pipe_wq_tasklet);
+ }
+ tasklet_kill(>pframe_pipe_wq_tasklet);
+#endif
+ wake_up_interruptible_nr(>pframe_wq, 1);
+
+ return 0;
+}
+
+static int
+gtp_gdbrsp_qtinit(void)
+{
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qtinit\n");
+#endif
+
+ if (gtp_start)
+ gtp_gdbrsp_qtstop();
+
+ gtp_list_release();
+
+ if (gtp_frame)
+ gtp_frame_reset();
+
+ gtpro_list_clear();
+
+ gtp_var_release();
+
+ return 0;
+}
+
+struct gtp_x_goto {
+ struct gtp_x_goto *next;
+ unsigned int addr;
+ int non_goto_done;
+};
+
+static struct gtp_x_goto *
+gtp_x_goto_find(struct gtp_x_goto *list, unsigned int pc)
+{
+ struct gtp_x_goto *ret = NULL;
+
+ for (ret = list; ret; ret = ret->next) {
+ if (ret->addr == pc)
+ break;
+ }
+
+ return ret;
+}
+
+static struct gtp_x_goto *
+gtp_x_goto_add(struct gtp_x_goto **list, unsigned int pc, int non_goto_done)
+{
+ struct gtp_x_goto *ret;
+
+ ret = kmalloc(sizeof(struct gtp_x_goto), GFP_KERNEL);
+ if (!ret)
+ goto out;
+
+ ret->addr = pc;
+ ret->non_goto_done = non_goto_done;
+
+ if (*list) {
+ ret->next = *list;
+ *list = ret;
+ } else {
+ ret->next = NULL;
+ *list = ret;
+ }
+
+out:
+ return ret;
+}
+
+struct gtp_x_var {
+ struct gtp_x_var *next;
+ unsigned int num;
+ unsigned int flags;
+};
+
+static int
+gtp_x_var_add(struct gtp_x_var **list, unsigned int num, unsigned int flag)
+{
+ struct gtp_x_var *curv;
+
+ for (curv = *list; curv; curv = curv->next) {
+ if (curv->num == num)
+ break;
+ }
+
+ if (!curv) {
+ curv = kmalloc(sizeof(struct gtp_x_var), GFP_KERNEL);
+ if (!curv)
+ return -ENOMEM;
+ curv->num = num;
+ curv->flags = 0;
+ if (*list) {
+ curv->next = *list;
+ *list = curv;
+ } else {
+ curv->next = NULL;
+ *list = curv;
+ }
+ }
+
+ curv->flags |= flag;
+
+ return 0;
+}
+
+static int
+gtp_check_x(struct gtp_entry *tpe, struct action *ae)
+{
+ int ret = -EINVAL;
+ unsigned int pc = 0, sp = 0;
+ struct gtp_x_goto *glist = NULL, *gtmp;
+ struct gtp_x_var *vlist = NULL, *vtmp;
+ uint8_t *ebuf = ae->u.exp.buf;
+ int last_trace_pc = -1;
+ unsigned int stack_size = 0;
+
+reswitch:
+ while (pc < ae->u.exp.size) {
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_check_x: cmd %x\n", ebuf[pc]);
+#endif
+ switch (ebuf[pc++]) {
+ /* add */
+ case 0x02:
+ /* sub */
+ case 0x03:
+ /* mul */
+ case 0x04:
+ /* lsh */
+ case 0x09:
+ /* rsh_signed */
+ case 0x0a:
+ /* rsh_unsigned */
+ case 0x0b:
+ /* bit_and */
+ case 0x0f:
+ /* bit_or */
+ case 0x10:
+ /* bit_xor */
+ case 0x11:
+ /* equal */
+ case 0x13:
+ /* less_signed */
+ case 0x14:
+ /* less_unsigned */
+ case 0x15:
+ /* pop */
+ case 0x29:
+ /* swap */
+ case 0x2b:
+ if (ae->type == 'X') {
+ if (sp < 1) {
+ printk(KERN_WARNING "gtp_check_x: "
+ "stack overflow "
+ "in %d.\n",
+ pc - 1);
+ goto release_out;
+ }
+ }
+ break;
+
+ /* trace */
+ case 0x0c:
+ if (tpe->have_printk)
+ last_trace_pc = pc - 1;
+
+ if (ae->type == 'X') {
+ if (sp < 2) {
+ printk(KERN_WARNING "gtp_check_x: "
+ "stack overflow "
+ "in %d.\n",
+ pc - 1);
+ goto release_out;
+ } else
+ sp -= 2;
+ }
+ break;
+
+ /* log_not */
+ case 0x0e:
+ /* bit_not */
+ case 0x12:
+ /* ref8 */
+ case 0x17:
+ /* ref16 */
+ case 0x18:
+ /* ref32 */
+ case 0x19:
+ /* ref64 */
+ case 0x1a:
+ break;
+
+ /* dup */
+ case 0x28:
+ if (ae->type == 'X') {
+ sp++;
+ if (stack_size < sp)
+ stack_size = sp;
+ }
+ break;
+
+ /* const8 */
+ case 0x22:
+ if (ae->type == 'X') {
+ sp++;
+ if (stack_size < sp)
+ stack_size = sp;
+ }
+ /* ext */
+ case 0x16:
+ /* zero_ext */
+ case 0x2a:
+ if (pc >= ae->u.exp.size)
+ goto release_out;
+ pc++;
+ break;
+
+ /* trace_quick */
+ case 0x0d:
+ if (tpe->have_printk)
+ last_trace_pc = pc - 1;
+
+ if (pc >= ae->u.exp.size)
+ goto release_out;
+ pc++;
+ break;
+
+ /* const16 */
+ case 0x23:
+ /* reg */
+ case 0x26:
+ if (pc + 1 >= ae->u.exp.size)
+ goto release_out;
+ pc += 2;
+ if (ae->type == 'X') {
+ sp++;
+ if (stack_size < sp)
+ stack_size = sp;
+ }
+ break;
+
+ /* const32 */
+ case 0x24:
+ if (pc + 3 >= ae->u.exp.size)
+ goto release_out;
+ pc += 4;
+ if (ae->type == 'X') {
+ sp++;
+ if (stack_size < sp)
+ stack_size = sp;
+ }
+ break;
+
+ /* const64 */
+ case 0x25:
+ if (pc + 7 >= ae->u.exp.size)
+ goto release_out;
+ pc += 8;
+ if (ae->type == 'X') {
+ sp++;
+ if (stack_size < sp)
+ stack_size = sp;
+ }
+ break;
+
+ /* if_goto */
+ case 0x20:
+ if (tpe->have_printk)
+ goto release_out;
+
+ if (pc + 1 >= ae->u.exp.size)
+ goto release_out;
+ gtmp = gtp_x_goto_find(glist, pc);
+ if (gtmp) {
+ if (gtmp->non_goto_done)
+ goto out;
+ else {
+ gtmp->non_goto_done = 1;
+ pc += 2;
+ }
+ } else {
+ if (!gtp_x_goto_add(&glist, pc, 0)) {
+ ret = -ENOMEM;
+ goto release_out;
+ }
+ pc = (ebuf[pc] << 8)
+ + (ebuf[pc + 1]);
+ }
+ /* Mark this action X need sp check when it exec. */
+ ae->type = 0xff;
+ break;
+
+ /* goto */
+ case 0x21:
+ if (pc + 1 >= ae->u.exp.size)
+ goto release_out;
+ gtmp = gtp_x_goto_find(glist, pc);
+ if (gtmp)
+ goto out;
+ else {
+ if (!gtp_x_goto_add(&glist, pc, 1)) {
+ ret = -ENOMEM;
+ goto release_out;
+ }
+ pc = (ebuf[pc] << 8) + (ebuf[pc + 1]);
+ }
+ break;
+
+ /* end */
+ case 0x27:
+ goto out;
+ break;
+
+ /* getv */
+ case 0x2c: {
+ int arg;
+
+ if (pc + 1 >= ae->u.exp.size)
+ goto release_out;
+ arg = ebuf[pc++];
+ arg = (arg << 8) + ebuf[pc++];
+
+ if (arg == GTP_VAR_NO_SELF_TRACE_ID) {
+ tpe->no_self_trace = 1;
+ ret = 1;
+ goto release_out;
+ }
+
+ if (GTP_VAR_NOT_GETV(arg)) {
+ printk(KERN_WARNING
+ "gtp_check_x: The tv %d cannot "
+ "get.\n", arg);
+ goto release_out;
+ }
+
+ if (!GTP_VAR_IS_SPECIAL(arg)) {
+ if (gtp_x_var_add(&vlist, arg, 1)) {
+ ret = -ENOMEM;
+ goto release_out;
+ }
+ }
+ }
+ if (ae->type == 'X') {
+ sp++;
+ if (stack_size < sp)
+ stack_size = sp;
+ }
+ break;
+
+ /* setv */
+ case 0x2d: {
+ int arg;
+
+ if (pc + 1 >= ae->u.exp.size)
+ goto release_out;
+ arg = ebuf[pc++];
+ arg = (arg << 8) + ebuf[pc++];
+
+ if (arg == GTP_VAR_NO_SELF_TRACE_ID) {
+ tpe->no_self_trace = 1;
+ ret = 1;
+ goto release_out;
+ } else if (arg == GTP_VAR_KRET_ID) {
+ /* XXX: still not set it value to
+ maxactive. */
+ tpe->is_kretprobe = 1;
+ ret = 1;
+ goto release_out;
+ }
+
+
+ if (GTP_VAR_NOT_SETV(arg)) {
+ printk(KERN_WARNING
+ "gtp_check_x: The tv %d cannot "
+ "set.\n", arg);
+ goto release_out;
+ }
+
+ if (arg == GTP_VAR_PRINTK_LEVEL_ID)
+ tpe->have_printk = 1;
+
+ if (!GTP_VAR_IS_SPECIAL(arg)) {
+ if (gtp_x_var_add(&vlist, arg, 2)) {
+ ret = -ENOMEM;
+ goto release_out;
+ }
+ }
+ }
+ break;
+
+ /* tracev */
+ case 0x2e: {
+ int arg;
+
+ if (tpe->have_printk)
+ last_trace_pc = pc - 1;
+
+ if (pc + 1 >= ae->u.exp.size)
+ goto release_out;
+ arg = ebuf[pc++];
+ arg = (arg << 8) + ebuf[pc++];
+
+ if (arg == GTP_VAR_NO_SELF_TRACE_ID) {
+ tpe->no_self_trace = 1;
+ ret = 1;
+ goto release_out;
+ }
+
+ if (GTP_VAR_NOT_TRACEV(arg)) {
+ printk(KERN_WARNING
+ "gtp_check_x: The tv %d cannot "
+ "trace.\n", arg);
+ goto release_out;
+ }
+
+ if (!GTP_VAR_IS_SPECIAL(arg)) {
+ if (gtp_x_var_add(&vlist, arg, 4)) {
+ ret = -ENOMEM;
+ goto release_out;
+ }
+ }
+ }
+ break;
+
+ /* printf */
+ case 0x31: {
+ int arg = ebuf[pc++];
+ if (arg && ae->type == 'X') {
+ if (sp < 1) {
+ printk(KERN_WARNING
+ "gtp_check_x: stack "
+ "overflow in %d.\n",
+ pc - 2);
+ goto release_out;
+ } else
+ sp--;
+ }
+ pc += strlen((char *)ebuf + pc) + 1;
+ }
+ break;
+
+ /* div_signed */
+ case 0x05:
+ /* div_unsigned */
+ case 0x06:
+ /* rem_signed */
+ case 0x07:
+ /* rem_unsigned */
+ case 0x08:
+#ifdef CONFIG_MIPS
+ /* XXX, mips don't have 64 bit div. */
+ goto release_out;
+#endif
+ if (ae->type == 'X') {
+ if (sp < 1) {
+ printk(KERN_WARNING "gtp_check_x: "
+ "stack overflow "
+ "in %d.\n",
+ pc - 1);
+ goto release_out;
+ } else
+ sp--;
+ }
+ break;
+
+ /* float */
+ case 0x01:
+ /* ref_float */
+ case 0x1b:
+ /* ref_double */
+ case 0x1c:
+ /* ref_long_double */
+ case 0x1d:
+ /* l_to_d */
+ case 0x1e:
+ /* d_to_l */
+ case 0x1f:
+ /* trace16 */
+ case 0x30:
+ default:
+ goto release_out;
+ break;
+ }
+ }
+ goto release_out;
+
+out:
+ for (gtmp = glist; gtmp; gtmp = gtmp->next) {
+ if (!gtmp->non_goto_done)
+ break;
+ }
+ if (gtmp) {
+ pc = gtmp->addr + 2;
+ gtmp->non_goto_done = 1;
+ goto reswitch;
+ }
+ if (stack_size >= STACK_MAX) {
+ printk(KERN_WARNING "gtp_check_x: stack overflow.");
+ goto release_out;
+ }
+ ret = 0;
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_check_x: Code is OK. sp_checked is %d. "
+ "stack_size is %d.\n",
+ (ae->type == 'X'), stack_size);
+#endif
+
+release_out:
+ while (glist) {
+ gtmp = glist;
+ glist = glist->next;
+ kfree(gtmp);
+ }
+ while (vlist) {
+ struct gtp_var *var;
+
+ vtmp = vlist;
+ vlist = vlist->next;
+
+ /* Get the var of vtmp. */
+ var = gtp_var_find(vtmp->num);
+ if (var == NULL) {
+ printk(KERN_WARNING "gtp_check_x: cannot find "
+ "tvar %d.\n", vtmp->num);
+ ret = -EINVAL;
+ } else {
+ if (var->per_cpu == NULL) {
+ if ((vtmp->flags & 2)
+ && ((vtmp->flags & 1) || (vtmp->flags & 4)))
+ ae->u.exp.need_var_lock = 1;
+ }
+ }
+ kfree(vtmp);
+ }
+
+ if (tpe->have_printk && last_trace_pc > -1) {
+ /* Set the last trace code to printk code. */
+ switch (ebuf[last_trace_pc]) {
+ /* trace */
+ case 0x0c:
+ ebuf[last_trace_pc] = 0xfd;
+ break;
+ /* trace_quick */
+ case 0x0d:
+ ebuf[last_trace_pc] = 0xfe;
+ break;
+ /* tracev */
+ case 0x2e:
+ ebuf[last_trace_pc] = 0xff;
+ break;
+ }
+ }
+
+ return ret;
+}
+
+#ifndef GTP_FRAME_SIMPLE
+static void
+gtpframe_pipe_wq_wake_up(unsigned long data)
+{
+ wake_up_interruptible_nr(>pframe_pipe_wq, 1);
+}
+#endif
+
+static void
+gtp_wq_add_work(unsigned long data)
+{
+ queue_work(gtp_wq, (struct work_struct *)data);
+}
+
+static int
+gtp_gdbrsp_qtstart(void)
+{
+ int cpu;
+ struct gtp_entry *tpe;
+ struct gtp_var *tve;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qtstart\n");
+#endif
+
+ if (gtp_start)
+ return -EBUSY;
+
+#ifndef GTP_FRAME_SIMPLE
+ if (!tracing_is_on()) {
+ printk(KERN_WARNING "qtstart: Ring buffer is off. Please use "
+ "command "
+ "\"echo 1 > /sys/kernel/debug/tracing/tracing_on\" "
+ "open it.\n");
+ return -EIO;
+ }
+#endif
+
+ for (tpe = gtp_list; tpe; tpe = tpe->next) {
+ int ret;
+ struct action *ae, *prev_ae = NULL;
+
+ /* Check cond. */
+ if (tpe->cond) {
+ ret = gtp_check_x(tpe, tpe->cond);
+ if (ret > 0) {
+ kfree(tpe->cond->u.exp.buf);
+ kfree(tpe->cond);
+ tpe->cond = NULL;
+ } else if (ret < 0)
+ return ret;
+ }
+
+ /* Check X. */
+ for (ae = tpe->action_list; ae; ae = ae->next) {
+re_check:
+ if (ae->type == 'X' || ae->type == 0xff) {
+ ret = gtp_check_x(tpe, ae);
+ if (ret > 0) {
+ struct action *old_ae = ae;
+
+ /* Remove ae from action_list. */
+ ae = ae->next;
+ if (prev_ae)
+ prev_ae->next = ae;
+ else
+ tpe->action_list = ae;
+
+ kfree(old_ae->u.exp.buf);
+ kfree(old_ae);
+
+ if (ae)
+ goto re_check;
+ else
+ break;
+ } else if (ret < 0)
+ return ret;
+ }
+
+ prev_ae = ae;
+ }
+
+ /* Check the tracepoint that have printk. */
+ if (tpe->have_printk) {
+ struct action *ae, *prev_ae = NULL;
+ struct gtpsrc *src, *srctail = NULL;
+
+restart:
+ for (ae = tpe->action_list; ae;
+ prev_ae = ae, ae = ae->next) {
+ switch (ae->type) {
+ case 'R':
+ /* Remove it. */
+ if (prev_ae)
+ prev_ae->next = ae->next;
+ else
+ tpe->action_list = ae->next;
+ kfree(ae->src);
+ kfree(ae);
+ if (prev_ae)
+ ae = prev_ae;
+ else
+ goto restart;
+ break;
+ case 'M':
+ printk(KERN_WARNING "qtstart: action "
+ "of tp %d %p is not right. "
+ "Please put global variable to "
+ "trace state variable "
+ "$printk_tmp before print it.\n",
+ (int)tpe->num,
+ (void *)(CORE_ADDR)tpe->addr);
+ return -EINVAL;
+ break;
+ }
+ }
+
+ for (src = tpe->src; src; src = src->next) {
+ int i;
+ char str[strlen(src->src) >> 1];
+ char *var = NULL;
+ ULONGEST num;
+ char tmp[30];
+ struct gtpsrc *ksrc;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qtstart: action "
+ "%s\n", src->src);
+#endif
+ /* Get the action in str. */
+ if (strncmp("cmd:0:", src->src,
+ strlen("cmd:0:")))
+ continue;
+ var = hex2ulongest(src->src + 6, &num);
+ if (var[0] == '\0')
+ return -EINVAL;
+ var++;
+ hex2string(var, str);
+ if (strlen(str) != num)
+ return -EINVAL;
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qtstart: action "
+ "command %s\n", str);
+#endif
+
+ if (strncmp("collect ", str,
+ strlen("collect ")))
+ continue;
+ for (i = strlen("collect "); ; i++) {
+ if (str[i] != ' ') {
+ var = str + i;
+ break;
+ }
+ if (str[i] == '\0')
+ break;
+ }
+ if (!var) {
+ printk(KERN_WARNING "qtstart: cannot "
+ "get the var name "
+ "from tp %d %p"
+ "command %s.\n",
+ (int)tpe->num,
+ (void *)(CORE_ADDR)tpe->addr,
+ str);
+ return -EINVAL;
+ }
+ if (strcmp(var, "$args") == 0
+ || strcmp(var, "$local") == 0) {
+ printk(KERN_WARNING "qtstart: cannot "
+ "print $args and "
+ "$local.\n");
+ return -EINVAL;
+ }
+ if (strcmp(var, "$reg") == 0)
+ continue;
+
+ ksrc = kmalloc(sizeof(struct gtpsrc),
+ GFP_KERNEL);
+ if (ksrc == NULL)
+ return -ENOMEM;
+ ksrc->next = NULL;
+
+ snprintf(tmp, 30, "gtp %d %p:", (int)tpe->num,
+ (void *)(CORE_ADDR)tpe->addr);
+ ksrc->src = kmalloc(strlen(tmp)
+ + strlen(var) + 2,
+ GFP_KERNEL);
+ if (ksrc->src == NULL) {
+ kfree(ksrc);
+ return -ENOMEM;
+ }
+ sprintf(ksrc->src, "%s%s=", tmp, var);
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qtstart: new "
+ "printk var %s\n", ksrc->src);
+#endif
+
+ if (tpe->printk_str)
+ srctail->next = ksrc;
+ else
+ tpe->printk_str = ksrc;
+ srctail = ksrc;
+ }
+ }
+ }
+
+ if (!gtp_frame) {
+#ifdef GTP_FRAME_SIMPLE
+ gtp_frame = vmalloc(GTP_FRAME_SIZE);
+#else
+ gtp_frame = ring_buffer_alloc(GTP_FRAME_SIZE,
+ gtp_circular ? RB_FL_OVERWRITE
+ : 0);
+#endif
+ if (!gtp_frame)
+ return -ENOMEM;
+
+ gtp_frame_reset();
+ }
+
+ for_each_online_cpu(cpu) {
+ per_cpu(rdtsc_current, cpu) = 0;
+ per_cpu(local_clock_current, cpu) = 0;
+ per_cpu(rdtsc_offset, cpu) = 0;
+ per_cpu(local_clock_offset, cpu) = 0;
+ }
+
+ gtp_start = 1;
+
+ gtp_var_array = kmalloc(sizeof(struct gtp_var *)
+ * (gtp_var_tail - gtp_var_head + 1),
+ GFP_KERNEL);
+ if (!gtp_var_array) {
+ gtp_gdbrsp_qtstop();
+ return -ENOMEM;
+ }
+ memset(gtp_var_array, '\0', sizeof(struct gtp_var *)
+ *(gtp_var_tail - gtp_var_head + 1));
+ for (tve = gtp_var_list; tve; tve = tve->next)
+ gtp_var_array[tve->num - gtp_var_head] = tve;
+
+#ifdef CONFIG_PERF_EVENTS
+ /* Clear pc_pe_list. */
+ for_each_online_cpu(cpu) {
+ per_cpu(pc_pe_list, cpu) = NULL;
+ per_cpu(pc_pe_list_all_disabled, cpu) = 1;
+ }
+ for (tve = gtp_var_list; tve; tve = tve->next) {
+ if (tve->ptid == pe_tv_unknown)
+ continue;
+ if (tve->pts->event)
+ continue;
+
+ /* Get event. */
+ tve->pts->event =
+ perf_event_create_kernel_counter(&(tve->pts->attr),
+ tve->pts->cpu,
+ NULL, NULL, NULL);
+ if (IS_ERR(tve->pts->event)) {
+ int ret = PTR_ERR(tve->pts->event);
+
+ printk(KERN_WARNING "gtp_gdbrsp_qtstart:"
+ "create perf_event %d got error.\n",
+ (int)tve->num);
+ tve->pts->event = NULL;
+ gtp_gdbrsp_qtstop();
+ return ret;
+ }
+
+ /* Add event to pc_pe_list. */
+ if (tve->pts->cpu >= 0) {
+ struct pe_tv_s *ppl = per_cpu(pc_pe_list,
+ tve->pts->cpu);
+ if (ppl == NULL) {
+ per_cpu(pc_pe_list, tve->pts->cpu) = tve->pts;
+ tve->pts->pc_next = NULL;
+ } else {
+ tve->pts->pc_next = ppl;
+ per_cpu(pc_pe_list,
+ tve->pts->cpu) = tve->pts;
+ }
+ if (tve->pts->en)
+ per_cpu(pc_pe_list_all_disabled, tve->pts->cpu)
+ = 0;
+ }
+ }
+#endif
+
+#ifndef GTP_FRAME_SIMPLE
+ tasklet_init(>pframe_pipe_wq_tasklet, gtpframe_pipe_wq_wake_up, 0);
+#endif
+
+ gtp_start_last_errno = 0;
+
+ for (tpe = gtp_list; tpe; tpe = tpe->next) {
+ tpe->reason = gtp_stop_normal;
+ if (!tpe->disable && tpe->addr != (CORE_ADDR)register_kprobe) {
+ int ret;
+
+ if (!tpe->nopass)
+ atomic_set(&tpe->current_pass, tpe->pass);
+
+ tasklet_init(&tpe->tasklet, gtp_wq_add_work,
+ (unsigned long)&tpe->work);
+
+ if (tpe->is_kretprobe) {
+ tpe->kp.handler = gtp_kp_ret_handler;
+ ret = register_kretprobe(&tpe->kp);
+ } else {
+ tpe->kp.kp.pre_handler = gtp_kp_pre_handler;
+ tpe->kp.kp.post_handler = gtp_kp_post_handler;
+ ret = register_kprobe(&tpe->kp.kp);
+ }
+ if (ret < 0) {
+ printk(KERN_WARNING "gtp_gdbrsp_qtstart:"
+ "register tracepoint %d %p got error.\n",
+ (int)tpe->num, (void *)(CORE_ADDR)tpe->addr);
+ if (gtp_start_ignore_error) {
+ gtp_start_last_errno = (uint64_t)ret;
+ continue;
+ } else {
+ gtp_gdbrsp_qtstop();
+ return ret;
+ }
+ }
+ tpe->kpreg = 1;
+ }
+ }
+
+ return 0;
+}
+
+static int
+gtp_parse_x(struct gtp_entry *tpe, struct action *ae, char **pkgp)
+{
+ ULONGEST size;
+ int ret = 0, i, h, l;
+ char *pkg = *pkgp;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_parse_x: %s\n", pkg);
+#endif
+
+ if (pkg[0] == '\0') {
+ ret = -EINVAL;
+ goto out;
+ }
+ pkg = hex2ulongest(pkg, &size);
+ if (pkg[0] != ',') {
+ ret = -EINVAL;
+ goto out;
+ }
+ ae->u.exp.size = (unsigned int)size;
+ pkg++;
+
+ ae->u.exp.buf = kmalloc(ae->u.exp.size, GFP_KERNEL);
+ if (!ae->u.exp.buf)
+ return -ENOMEM;
+
+ for (i = 0; i < ae->u.exp.size
+ && hex2int(pkg[0], &h) && hex2int(pkg[1], &l);
+ i++) {
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_parse_x: %s %d %d\n", pkg, h, l);
+#endif
+ ae->u.exp.buf[i] = (h << 4) | l;
+ pkg += 2;
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_parse_x: %x\n", ae->u.exp.buf[i]);
+#endif
+ }
+ if (i != ae->u.exp.size) {
+ kfree(ae->u.exp.buf);
+ ret = -EINVAL;
+ goto out;
+ }
+
+ ae->u.exp.need_var_lock = 0;
+
+out:
+ *pkgp = pkg;
+ return ret;
+}
+
+static int
+gtp_gdbrsp_qtdp(char *pkg)
+{
+ int addnew = 1;
+ ULONGEST num, addr;
+ struct gtp_entry *tpe;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qtdp: %s\n", pkg);
+#endif
+
+ if (gtp_start)
+ return -EBUSY;
+
+ if (pkg[0] == '-') {
+ pkg++;
+ addnew = 0;
+ }
+
+ /* Get num and addr. */
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg = hex2ulongest(pkg, &num);
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg++;
+ pkg = hex2ulongest(pkg, &addr);
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg++;
+
+ tpe = gtp_list_find(num, addr);
+ if (addnew) {
+ if (tpe)
+ return -EINVAL;
+
+ tpe = gtp_list_add(num, addr);
+ if (tpe == NULL)
+ return -ENOMEM;
+
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ if (pkg[0] == 'D')
+ tpe->disable = 1;
+ pkg++;
+
+ /* Get step and pass. */
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg++;
+ pkg = hex2ulongest(pkg, &tpe->step);
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg++;
+ pkg = hex2ulongest(pkg, &tpe->pass);
+ if (tpe->pass == 0)
+ tpe->nopass = 1;
+ }
+
+ if (tpe) {
+ /* Add action to tpe. */
+ int step_action = 0;
+
+ if (pkg[0] == 'S') {
+ pkg++;
+ step_action = 1;
+ /* XXX: Still not support step. */
+ return 1;
+ }
+ while (pkg[0]) {
+ struct action *ae = NULL, *atail = NULL;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qtdp: %s\n", pkg);
+#endif
+ switch (pkg[0]) {
+ case ':':
+ pkg++;
+ break;
+ case 'M': {
+ int is_neg = 0;
+ ULONGEST ulongtmp;
+
+ ae = gtp_action_alloc(pkg);
+ if (!ae)
+ return -ENOMEM;
+ pkg++;
+ if (pkg[0] == '-') {
+ is_neg = 1;
+ pkg++;
+ }
+ pkg = hex2ulongest(pkg, &ulongtmp);
+ ae->u.m.regnum = (int)ulongtmp;
+ if (is_neg)
+ ae->u.m.regnum
+ = -ae->u.m.regnum;
+ if (pkg[0] == '\0') {
+ kfree(ae);
+ return -EINVAL;
+ }
+ pkg++;
+ pkg = hex2ulongest(pkg, &ulongtmp);
+ ae->u.m.offset = (CORE_ADDR)ulongtmp;
+ if (pkg[0] == '\0') {
+ kfree(ae);
+ return -EINVAL;
+ }
+ pkg++;
+ pkg = hex2ulongest(pkg, &ulongtmp);
+ ae->u.m.size = (size_t)ulongtmp;
+ }
+ break;
+ case 'R':
+ /* XXX: reg_mask is ignore. */
+ ae = gtp_action_alloc(pkg);
+ if (!ae)
+ return -ENOMEM;
+ pkg++;
+ pkg = hex2ulongest(pkg,
+ &ae->u.reg_mask);
+ break;
+ case 'X': {
+ int ret;
+
+ ae = gtp_action_alloc(pkg);
+ if (!ae)
+ return -ENOMEM;
+ pkg++;
+ ret = gtp_parse_x(tpe, ae, &pkg);
+ if (ret) {
+ kfree(ae);
+ ae = NULL;
+
+ if (ret < 0)
+ return ret;
+ }
+ }
+ break;
+ case '-':
+ pkg++;
+ break;
+ default:
+ /* XXX: Not support. */
+ return 1;
+ }
+
+ if (ae) {
+ /* Save the src. */
+ ae->src = gtp_strdup(ae->src, pkg);
+ if (ae->src == NULL) {
+ kfree(ae);
+ return -ENOMEM;
+ }
+ /* Add ae to tpe. */
+ if ((ae->type == 'X' || ae->type == 0xff)
+ && addnew && !tpe->cond) {
+ tpe->cond = ae;
+ tpe->cond->next = NULL;
+ } else if (!tpe->action_list) {
+ tpe->action_list = ae;
+ atail = ae;
+ } else {
+ if (atail == NULL)
+ for (atail = tpe->action_list;
+ atail->next;
+ atail = atail->next)
+ ;
+ atail->next = ae;
+ atail = ae;
+ }
+ }
+ }
+ } else
+ return -EINVAL;
+
+ return 0;
+}
+
+static int
+gtp_gdbrsp_qtdpsrc(char *pkg)
+{
+ ULONGEST num, addr;
+ struct gtpsrc *src, *srctail;
+ struct gtp_entry *tpe;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qtdpsrc: %s\n", pkg);
+#endif
+
+ if (gtp_start)
+ return -EBUSY;
+
+ /* Get num and addr. */
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg = hex2ulongest(pkg, &num);
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg++;
+ pkg = hex2ulongest(pkg, &addr);
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg++;
+ tpe = gtp_list_find(num, addr);
+ if (tpe == NULL)
+ return -EINVAL;
+
+ src = kmalloc(sizeof(struct gtpsrc), GFP_KERNEL);
+ if (src == NULL)
+ return -ENOMEM;
+ src->next = NULL;
+ src->src = gtp_strdup(pkg, NULL);
+ if (src->src == NULL) {
+ kfree(src);
+ return -ENOMEM;
+ }
+
+ if (tpe->src) {
+ for (srctail = tpe->src; srctail->next;
+ srctail = srctail->next)
+ ;
+ srctail->next = src;
+ } else
+ tpe->src = src;
+
+ return 0;
+}
+
+static int
+gtp_gdbrsp_qtdisconnected(char *pkg)
+{
+ ULONGEST setting;
+
+ if (pkg[0] == '\0')
+ return -EINVAL;
+
+ hex2ulongest(pkg, &setting);
+ gtp_disconnected_tracing = (int)setting;
+
+ return 0;
+}
+
+static int
+gtp_gdbrsp_qtbuffer(char *pkg)
+{
+ if (strncmp("circular:", pkg, 9) == 0) {
+ ULONGEST setting;
+
+ pkg += 9;
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ hex2ulongest(pkg, &setting);
+
+ gtp_circular = (int)setting;
+ if (gtp_frame)
+ ring_buffer_change_overwrite(gtp_frame, (int)setting);
+
+ return 0;
+ }
+
+ return 1;
+}
+
+static int
+gtp_frame_head_find_num(int num)
+{
+#ifdef GTP_FRAME_SIMPLE
+ int tfnum = 0;
+ char *tmp = gtp_frame_r_start;
+
+ do {
+ if (tmp == gtp_frame_end)
+ tmp = gtp_frame;
+
+ if (FID(tmp) == FID_HEAD) {
+ if (tfnum == num) {
+ gtp_frame_current_num = num;
+ gtp_frame_current = tmp;
+ return 0;
+ }
+ tfnum++;
+ }
+
+ tmp = gtp_frame_next(tmp);
+ if (!tmp)
+ break;
+ } while (tmp != gtp_frame_w_start);
+#else
+ if (gtp_frame_current_num >= num)
+ gtp_frame_iter_reset();
+
+ while (1) {
+ int cpu;
+
+ cpu = gtp_frame_iter_peek_head();
+ if (cpu < 0)
+ break;
+
+ if (num == gtp_frame_current_num)
+ return cpu;
+
+ ring_buffer_read(gtp_frame_iter[cpu], NULL);
+ }
+#endif
+
+ return -1;
+}
+
+static int
+gtp_frame_head_find_addr(int inside, unsigned long lo,
+ unsigned long hi)
+{
+#ifdef GTP_FRAME_SIMPLE
+ int tfnum = gtp_frame_current_num;
+ char *tmp;
+
+ if (gtp_frame_current)
+ tmp = gtp_frame_current;
+ else
+ tmp = gtp_frame_r_start;
+
+ do {
+ if (tmp == gtp_frame_end)
+ tmp = gtp_frame;
+
+ if (FID(tmp) == FID_HEAD) {
+ if (tfnum != gtp_frame_current_num) {
+ char *next;
+ struct pt_regs *regs = NULL;
+
+ for (next = *(char **)(tmp + FID_SIZE); next;
+ next = *(char **)(next + FID_SIZE)) {
+ if (FID(next) == FID_REG) {
+ regs = (struct pt_regs *)
+ (next + FID_SIZE
+ + sizeof(char *));
+ break;
+ }
+ }
+ if (regs
+ && ((inside
+ && GTP_REGS_PC(regs) >= lo
+ && GTP_REGS_PC(regs) <= hi)
+ || (!inside
+ && (GTP_REGS_PC(regs) < lo
+ || GTP_REGS_PC(regs) > hi)))) {
+ gtp_frame_current_num = tfnum;
+ gtp_frame_current = tmp;
+ return 0;
+ }
+ }
+ tfnum++;
+ }
+
+ tmp = gtp_frame_next(tmp);
+ if (!tmp)
+ break;
+ } while (tmp != gtp_frame_w_start);
+#else
+ while (1) {
+ int cpu;
+ struct ring_buffer_event *rbe;
+ char *tmp;
+ struct pt_regs *regs = NULL;
+
+ cpu = gtp_frame_iter_peek_head();
+ if (cpu < 0)
+ break;
+
+ while (1) {
+ ring_buffer_read(gtp_frame_iter[cpu], NULL);
+ rbe = ring_buffer_iter_peek(gtp_frame_iter[cpu], NULL);
+ if (rbe == NULL)
+ break;
+
+ tmp = ring_buffer_event_data(rbe);
+ if (FID(tmp) == FID_HEAD)
+ break;
+ if (FID(tmp) == FID_REG) {
+ regs = (struct pt_regs *)(tmp + FID_SIZE);
+ break;
+ }
+ }
+
+ if (regs
+ && ((inside
+ && GTP_REGS_PC(regs) >= lo
+ && GTP_REGS_PC(regs) <= hi)
+ || (!inside
+ && (GTP_REGS_PC(regs) < lo
+ || GTP_REGS_PC(regs) > hi))))
+ return gtp_frame_head_find_num(gtp_frame_current_num);
+ }
+#endif
+
+ return -1;
+}
+
+static int
+gtp_frame_head_find_trace(ULONGEST trace)
+{
+#ifdef GTP_FRAME_SIMPLE
+ int tfnum = gtp_frame_current_num;
+ char *tmp;
+
+ if (gtp_frame_current)
+ tmp = gtp_frame_current;
+ else
+ tmp = gtp_frame_r_start;
+
+ do {
+ if (tmp == gtp_frame_end)
+ tmp = gtp_frame;
+
+ if (FID(tmp) == FID_HEAD) {
+ if (tfnum != gtp_frame_current_num) {
+ if (trace == *(ULONGEST *) (tmp + FID_SIZE
+ + sizeof(char *))) {
+ gtp_frame_current_num = tfnum;
+ gtp_frame_current = tmp;
+ return 0;
+ }
+ }
+ tfnum++;
+ }
+
+ tmp = gtp_frame_next(tmp);
+ if (!tmp)
+ break;
+ } while (tmp != gtp_frame_w_start);
+#else
+ while (1) {
+ int cpu;
+ struct ring_buffer_event *rbe;
+ char *tmp;
+
+ cpu = gtp_frame_iter_peek_head();
+ if (cpu < 0)
+ break;
+
+ rbe = ring_buffer_iter_peek(gtp_frame_iter[cpu], NULL);
+ if (rbe == NULL) {
+ /* It will not happen, just for safe. */
+ return -1;
+ }
+ tmp = ring_buffer_event_data(rbe);
+ if (trace == *(ULONGEST *) (tmp + FID_SIZE))
+ return cpu;
+
+ ring_buffer_read(gtp_frame_iter[cpu], NULL);
+ }
+#endif
+
+ return -1;
+}
+
+static int
+gtp_gdbrsp_qtframe(char *pkg)
+{
+ int ret = -1;
+#ifndef GTP_FRAME_SIMPLE
+ int old_num = gtp_frame_current_num;
+#endif
+
+ if (gtp_start)
+ return -EBUSY;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qtframe: %s\n", pkg);
+#endif
+
+ /* XXX: If this check is removed, add a check for gtp_frame. */
+ if (atomic_read(>p_frame_create) == 0)
+ goto out;
+
+ if (strncmp(pkg, "pc:", 3) == 0) {
+ ULONGEST addr;
+
+ pkg += 3;
+
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ hex2ulongest(pkg, &addr);
+
+ ret = gtp_frame_head_find_addr(1, (unsigned long)addr,
+ (unsigned long)addr);
+ } else if (strncmp(pkg, "tdp:", 4) == 0) {
+ ULONGEST trace;
+
+ pkg += 4;
+
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ hex2ulongest(pkg, &trace);
+
+ ret = gtp_frame_head_find_trace(trace);
+ } else if (strncmp(pkg, "range:", 6) == 0) {
+ ULONGEST start, end;
+
+ pkg += 6;
+
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg = hex2ulongest(pkg, &start);
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg++;
+ hex2ulongest(pkg, &end);
+
+ ret = gtp_frame_head_find_addr(1, (unsigned long)start,
+ (unsigned long)end);
+ } else if (strncmp(pkg, "outside:", 8) == 0) {
+ ULONGEST start, end;
+
+ pkg += 8;
+
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg = hex2ulongest(pkg, &start);
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg++;
+ hex2ulongest(pkg, &end);
+
+ ret = gtp_frame_head_find_addr(0, (unsigned long)start,
+ (unsigned long)end);
+ } else {
+ ULONGEST num;
+
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ hex2ulongest(pkg, &num);
+
+ if (((int) num) < 0) {
+ /* Return to current. */
+#ifdef GTP_FRAME_SIMPLE
+ gtp_frame_current = NULL;
+ gtp_frame_current_num = -1;
+#else
+ gtp_frame_iter_reset();
+#endif
+
+ return 0;
+ }
+ ret = gtp_frame_head_find_num((int) num);
+ }
+
+out:
+ if (ret < 0) {
+#ifndef GTP_FRAME_SIMPLE
+ /* Set frame back to old_num. */
+ if (old_num < 0)
+ gtp_frame_iter_reset();
+ else
+ gtp_frame_head_find_num(old_num);
+#endif
+ strcpy(gtp_rw_bufp, "F-1");
+ gtp_rw_bufp += 3;
+ gtp_rw_size += 3;
+ } else {
+#ifdef GTP_FRAME_SIMPLE
+ gtp_frame_current_tpe = *(ULONGEST *)(gtp_frame_current
+ + FID_SIZE
+ + sizeof(char *));
+ sprintf(gtp_rw_bufp, "F%xT%x", gtp_frame_current_num,
+ (unsigned int) gtp_frame_current_tpe);
+#else
+ struct ring_buffer_event *rbe;
+ char *tmp;
+
+ rbe = ring_buffer_read(gtp_frame_iter[ret],
+ >p_frame_current_clock);
+ if (rbe == NULL) {
+ /* It will not happen, just for safe. */
+ ret = -1;
+ goto out;
+ }
+ gtp_frame_current_cpu = ret;
+ tmp = ring_buffer_event_data(rbe);
+ gtp_frame_current_tpe = *(ULONGEST *)(tmp + FID_SIZE);
+ sprintf(gtp_rw_bufp, "F%xT%x", gtp_frame_current_num,
+ (unsigned int) gtp_frame_current_tpe);
+#endif
+ gtp_rw_size += strlen(gtp_rw_bufp);
+ gtp_rw_bufp += strlen(gtp_rw_bufp);
+ }
+ return 1;
+}
+
+static int
+gtp_gdbrsp_qtro(char *pkg)
+{
+ ULONGEST start, end;
+
+ gtpro_list_clear();
+
+ while (pkg[0]) {
+ pkg = hex2ulongest(pkg, &start);
+ if (pkg[0] != ',')
+ return -EINVAL;
+ pkg++;
+ pkg = hex2ulongest(pkg, &end);
+ if (pkg[0])
+ pkg++;
+
+ if (gtpro_list_add((CORE_ADDR)start, (CORE_ADDR)end) == NULL)
+ return -ENOMEM;
+ }
+
+ return 0;
+}
+
+static int
+gtp_gdbrsp_qtdv(char *pkg)
+{
+ ULONGEST num, val;
+ struct gtp_var *var;
+ char *src;
+ char *src_no_val;
+ int src_no_val_size;
+ int per_cpu_id = 0;
+ struct gtp_var **per_cpu = NULL;
+ int per_cpu_alloced = 0;
+ int ret = -EINVAL;
+#ifdef CONFIG_PERF_EVENTS
+ enum pe_tv_id ptid = pe_tv_unknown;
+ struct pe_tv_s *pts = NULL;
+ int pts_alloced = 0;
+#endif
+
+ pkg = hex2ulongest(pkg, &num);
+ if (GTP_VAR_IS_SPECIAL(num))
+ return 0;
+ if (pkg[0] != ':')
+ goto error_out;
+ pkg++;
+ src = pkg;
+ pkg = hex2ulongest(pkg, &val);
+ if (pkg[0] != ':')
+ goto error_out;
+
+ /* src_no_val is not include the val but the ':' after it. */
+ src_no_val = pkg;
+ src_no_val_size = strlen(src_no_val);
+
+ pkg++;
+
+ var = gtp_var_find(num);
+ if (var)
+ goto error_out;
+
+ /* Check if this is a "ignore_error" trace state variable. */
+ if (strcasecmp(pkg, "0:69676e6f72655f6572726f72") == 0)
+ gtp_start_ignore_error = 1;
+
+ /* Check if this is a "pc_" or "per_cpu_" trace state variable. */
+ if (strncasecmp(pkg, "0:70635f", 8) == 0
+ || strncasecmp(pkg, "0:7065725f6370755f", 18) == 0) {
+ int name_size;
+ char *id_s;
+ int mul = 1;
+ struct gtp_var *tve;
+
+ if (strncasecmp(pkg, "0:70635f", 8) == 0)
+ pkg += 8;
+ else
+ pkg += 18;
+ name_size = strlen(pkg);
+
+ /* Get the cpu id of this variable. */
+ if (name_size % 2 != 0)
+ goto error_out;
+ for (id_s = pkg + name_size - 2; id_s > pkg; id_s -= 2) {
+ int i, j;
+
+ if (!hex2int(id_s[0], &i))
+ goto error_out;
+ if (!hex2int(id_s[1], &j))
+ goto error_out;
+ j |= (i << 4);
+ if (j < 0x30 || j > 0x39)
+ break;
+ j -= 0x30;
+ per_cpu_id += mul * j;
+ mul *= 10;
+ /* src_no_val_size will not include the cpu id. */
+ src_no_val_size -= 2;
+ }
+ if (mul == 1)
+ goto error_out;
+ if (per_cpu_id >= gtp_cpu_number) {
+ printk(KERN_WARNING "gtp_gdbrsp_qtdv: id %d is bigger "
+ "than cpu number %d.\n",
+ per_cpu_id, gtp_cpu_number);
+ goto error_out;
+ }
+
+ /* Find the per cpu array per_cpu. */
+ for (tve = gtp_var_list; tve; tve = tve->next) {
+ if (tve->per_cpu) {
+ char *gtp_var_src;
+ /* Let gtp_var_src point after the value. */
+ gtp_var_src = hex2ulongest(tve->src, NULL);
+
+ if (strncmp(gtp_var_src, src_no_val,
+ src_no_val_size) == 0) {
+ per_cpu = tve->per_cpu;
+ break;
+ }
+ }
+ }
+ if (per_cpu == NULL) {
+ per_cpu = kcalloc(gtp_cpu_number,
+ sizeof(struct gtp_var *),
+ GFP_KERNEL);
+ if (per_cpu == NULL) {
+ ret = -ENOMEM;
+ goto error_out;
+ }
+ per_cpu_alloced = 1;
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qtdv: Create a "
+ "new per_cpu list for %s and set var "
+ "to cpu %d.\n",
+ src_no_val, per_cpu_id);
+#endif
+ } else {
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qtdv: Find a "
+ "per_cpu list for %s and set var "
+ "to cpu %d.\n",
+ src_no_val, per_cpu_id);
+#endif
+ }
+ } else {
+ /* Remove first "0:" for following code. */
+ if (strlen(pkg) <= 2)
+ goto error_out;
+ pkg += 2;
+ }
+
+ /* Check if this is a "pe_" OR "perf_event_" trace state variable. */
+ if (strncasecmp(pkg, "70655f", 6) == 0
+ || strncasecmp(pkg, "706572665f6576656e745f", 22) == 0) {
+#ifdef CONFIG_PERF_EVENTS
+ struct gtp_var *tve;
+
+ if (strncasecmp(pkg, "70655f", 6) == 0)
+ pkg += 6;
+ else
+ pkg += 22;
+
+ if (strncasecmp(pkg, "6370755f", 8) == 0) {
+ /* "cpu_" */
+ pkg += 8;
+ ptid = pe_tv_cpu;
+ } else if (strncasecmp(pkg, "747970655f", 10) == 0) {
+ /* "type_" */
+ pkg += 10;
+ ptid = pe_tv_type;
+ } else if (strncasecmp(pkg, "636f6e6669675f", 14) == 0) {
+ /* "config_" */
+ pkg += 14;
+ ptid = pe_tv_config;
+ } else if (strncasecmp(pkg, "656e5f", 6) == 0) {
+ /* "en_" */
+ pkg += 6;
+ ptid = pe_tv_en;
+ } else if (strncasecmp(pkg, "76616c5f", 8) == 0) {
+ /* "val_" */
+ pkg += 8;
+ ptid = pe_tv_val;
+ } else if (strncasecmp(pkg, "656e61626c65645f", 16) == 0) {
+ /* "enabled_" */
+ pkg += 16;
+ ptid = pe_tv_enabled;
+ } else if (strncasecmp(pkg, "72756e6e696e675f", 16) == 0) {
+ /* "running_" */
+ pkg += 16;
+ ptid = pe_tv_running;
+ } else
+ goto pe_format_error;
+
+ if (strlen(pkg) <= 0)
+ goto pe_format_error;
+
+ /* Find the pe_tv that name is pkg. */
+ for (tve = gtp_var_list; tve; tve = tve->next) {
+ if (tve->ptid != pe_tv_unknown) {
+ if (strcmp(tve->pts->name, pkg) == 0)
+ break;
+ }
+ }
+
+ if (tve)
+ pts = tve->pts;
+ else {
+ pts = kcalloc(1, sizeof(struct pe_tv_s), GFP_KERNEL);
+ if (pts == NULL) {
+ ret = -ENOMEM;
+ goto error_out;
+ }
+ pts_alloced = 1;
+ /* Init the value in pts to default value. */
+ pts->name = gtp_strdup(pkg, NULL);
+ if (per_cpu)
+ pts->cpu = per_cpu_id;
+ else
+ pts->cpu = -1;
+ pts->en = 0;
+ pts->attr.type = PERF_TYPE_HARDWARE;
+ pts->attr.config = PERF_COUNT_HW_CPU_CYCLES;
+ pts->attr.disabled = 1;
+ pts->attr.pinned = 1;
+ pts->attr.size = sizeof(struct perf_event_attr);
+ }
+
+ /* Set current val to pts. */
+ switch (ptid) {
+ case pe_tv_cpu:
+ pts->cpu = (int)(LONGEST)val;
+ break;
+ case pe_tv_type:
+ pts->attr.type = val;
+ break;
+ case pe_tv_config:
+ pts->attr.config = val;
+ break;
+ case pe_tv_en:
+ if (val) {
+ pts->attr.disabled = 0;
+ pts->en = 1;
+ } else {
+ pts->attr.disabled = 1;
+ pts->en = 0;
+ }
+ break;
+ case pe_tv_val:
+ case pe_tv_enabled:
+ case pe_tv_running:
+ break;
+ default:
+ goto pe_format_error;
+ break;
+ }
+#else
+ printk(KERN_WARNING "Current Kernel doesn't open "
+ "CONFIG_PERF_EVENTS\n");
+ ret = -ENXIO;
+ goto error_out;
+#endif
+ }
+
+#ifdef CONFIG_PERF_EVENTS
+ if (!gtp_var_add((unsigned int)num, (uint64_t)val, src,
+ per_cpu, per_cpu_id, ptid, pts)) {
+#else
+ if (!gtp_var_add((unsigned int)num, (uint64_t)val, src,
+ per_cpu, per_cpu_id)) {
+#endif
+ ret = -ENOMEM;
+ goto error_out;
+ }
+
+ return 0;
+
+#ifdef CONFIG_PERF_EVENTS
+pe_format_error:
+ printk(KERN_WARNING "The format of this perf event "
+ "trace state variables is not right.\n");
+#endif
+
+error_out:
+#ifdef CONFIG_PERF_EVENTS
+ if (pts_alloced)
+ kfree(pts);
+#endif
+ if (per_cpu_alloced)
+ kfree(per_cpu);
+ return ret;
+}
+
+static int
+gtp_gdbrsp_QT(char *pkg)
+{
+ int ret = 1;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_QT: %s\n", pkg);
+#endif
+
+ if (strcmp("init", pkg) == 0)
+ ret = gtp_gdbrsp_qtinit();
+ else if (strcmp("Stop", pkg) == 0)
+ ret = gtp_gdbrsp_qtstop();
+ else if (strcmp("Start", pkg) == 0)
+ ret = gtp_gdbrsp_qtstart();
+ else if (strncmp("DP:", pkg, 3) == 0)
+ ret = gtp_gdbrsp_qtdp(pkg + 3);
+ else if (strncmp("DPsrc:", pkg, 6) == 0)
+ ret = gtp_gdbrsp_qtdpsrc(pkg + 6);
+ else if (strncmp("Disconnected:", pkg, 13) == 0)
+ ret = gtp_gdbrsp_qtdisconnected(pkg + 13);
+ else if (strncmp("Buffer:", pkg, 7) == 0)
+ ret = gtp_gdbrsp_qtbuffer(pkg + 7);
+ else if (strncmp("Frame:", pkg, 6) == 0)
+ ret = gtp_gdbrsp_qtframe(pkg + 6);
+ else if (strncmp("ro:", pkg, 3) == 0)
+ ret = gtp_gdbrsp_qtro(pkg + 3);
+ else if (strncmp("DV:", pkg, 3) == 0)
+ ret = gtp_gdbrsp_qtdv(pkg + 3);
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_QT: return %d\n", ret);
+#endif
+
+ return ret;
+}
+
+static int
+gtp_get_status(struct gtp_entry *tpe, char *buf)
+{
+ int size = 0;
+ int tfnum = 0;
+ CORE_ADDR tmpaddr;
+
+ if (!gtp_frame) {
+ sprintf(buf, "tnotrun:0;");
+ buf += 10;
+ size += 10;
+ } else if (!tpe || (tpe && tpe->reason == gtp_stop_normal)) {
+ sprintf(buf, "tstop:0;");
+ buf += 8;
+ size += 8;
+ } else {
+ char outtmp[100];
+
+ switch (tpe->reason) {
+ case gtp_stop_frame_full:
+ sprintf(buf, "tfull:%lx;",
+ (unsigned long)tpe->num);
+ break;
+ case gtp_stop_efault:
+ sprintf(buf, "terror:%s:%lx;",
+ string2hex("read memory false", outtmp),
+ (unsigned long)tpe->num);
+ break;
+ case gtp_stop_access_wrong_reg:
+ sprintf(buf, "terror:%s:%lx;",
+ string2hex("access wrong register", outtmp),
+ (unsigned long)tpe->num);
+ break;
+ case gtp_stop_agent_expr_code_error:
+ sprintf(buf, "terror:%s:%lx;",
+ string2hex("agent expression code error",
+ outtmp),
+ (unsigned long)tpe->num);
+ break;
+ case gtp_stop_agent_expr_stack_overflow:
+ sprintf(buf, "terror:%s:%lx;",
+ string2hex("agent expression stack overflow",
+ outtmp),
+ (unsigned long)tpe->num);
+ break;
+ default:
+ buf[0] = '\0';
+ break;
+ }
+
+ size += strlen(buf);
+ buf += strlen(buf);
+ }
+
+ if (atomic_read(>p_frame_create)) {
+#ifdef GTP_FRAME_SIMPLE
+ char *tmp = gtp_frame_r_start;
+
+ do {
+ if (tmp == gtp_frame_end)
+ tmp = gtp_frame;
+
+ if (FID(tmp) == FID_HEAD)
+ tfnum++;
+
+ tmp = gtp_frame_next(tmp);
+ if (!tmp)
+ break;
+ } while (tmp != gtp_frame_w_start);
+#else
+ if (gtp_start) {
+ /* XXX: It is just the number of entries. */
+ tfnum = (int)ring_buffer_entries(gtp_frame);
+ } else {
+ int old_num = gtp_frame_current_num;
+ int cpu;
+
+ gtp_frame_iter_reset();
+
+ for_each_online_cpu(cpu) {
+ char *tmp;
+ struct ring_buffer_event *rbe;
+
+ while (1) {
+ rbe = ring_buffer_read
+ (gtp_frame_iter[cpu], NULL);
+ if (rbe == NULL)
+ break;
+ tmp = ring_buffer_event_data(rbe);
+ if (FID(tmp) == FID_HEAD)
+ tfnum++;
+ }
+ }
+
+ if (old_num == -1)
+ gtp_frame_iter_reset();
+ else if (old_num >= 0) {
+ gtp_frame_head_find_num(old_num);
+ ring_buffer_read
+ (gtp_frame_iter[gtp_frame_current_cpu],
+ NULL);
+ }
+ }
+#endif
+ }
+
+ sprintf(buf, "tframes:%x;", tfnum);
+ size += strlen(buf);
+ buf += strlen(buf);
+
+ sprintf(buf, "tcreated:%x;", atomic_read(>p_frame_create));
+ size += strlen(buf);
+ buf += strlen(buf);
+
+#ifdef GTP_FRAME_SIMPLE
+ sprintf(buf, "tsize:%x;", GTP_FRAME_SIZE);
+#else
+ if (gtp_frame)
+ sprintf(buf, "tsize:%lx;", ring_buffer_size(gtp_frame));
+ else
+ sprintf(buf, "tsize:%x;", GTP_FRAME_SIZE * num_online_cpus());
+#endif
+ size += strlen(buf);
+ buf += strlen(buf);
+
+#ifdef GTP_FRAME_SIMPLE
+ spin_lock(>p_frame_lock);
+ if (gtp_frame_is_circular)
+ tmpaddr = 0;
+ else
+ tmpaddr = GTP_FRAME_SIZE - (gtp_frame_w_start - gtp_frame);
+ spin_unlock(>p_frame_lock);
+#else
+ /* XXX: Ring buffer don't have interface to get the size of free
+ buffer. */
+ tmpaddr = 0;
+#endif
+ sprintf(buf, "tfree:%lx;", (unsigned long)tmpaddr);
+ size += strlen(buf);
+ buf += strlen(buf);
+
+ sprintf(buf, "circular:%x;", gtp_circular);
+ size += strlen(buf);
+ buf += strlen(buf);
+
+ sprintf(buf, "disconn:%x", gtp_disconnected_tracing);
+ size += strlen(buf);
+ buf += strlen(buf);
+
+ return size;
+}
+
+static int
+gtp_gdbrsp_qtstatus(void)
+{
+ struct gtp_entry *tpe;
+ int tmp;
+
+ for (tpe = gtp_list; tpe; tpe = tpe->next) {
+ if (tpe->reason != gtp_stop_normal)
+ break;
+ }
+
+ if (gtp_start && tpe) /* Tpe is stop, stop all tpes. */
+ gtp_gdbrsp_qtstop();
+
+ sprintf(gtp_rw_bufp, "T%x;", gtp_start ? 1 : 0);
+ gtp_rw_bufp += 3;
+ gtp_rw_size += 3;
+
+ tmp = gtp_get_status(tpe, gtp_rw_bufp);
+ gtp_rw_bufp += tmp;
+ gtp_rw_size += tmp;
+
+ return 1;
+}
+
+static void
+gtp_report_tracepoint(struct gtp_entry *gtp, char *buf)
+{
+ sprintf(buf, "T%lx:%lx:%c:%lx:%lx",
+ (unsigned long)gtp->num,
+ (unsigned long)gtp->addr,
+ (gtp->disable ? 'D' : 'E'),
+ (unsigned long)gtp->step,
+ (unsigned long)gtp->pass);
+}
+
+static void
+gtp_report_action(struct gtp_entry *gtp, struct action *action, char *buf)
+{
+ sprintf(buf, "A%lx:%lx:%s",
+ (unsigned long)gtp->num,
+ (unsigned long)gtp->addr,
+ action->src);
+}
+
+static void
+gtp_report_src(struct gtp_entry *gtp, struct gtpsrc *src, char *buf)
+{
+ sprintf(buf, "Z%lx:%lx:%s",
+ (unsigned long)gtp->num,
+ (unsigned long)gtp->addr,
+ src->src);
+}
+
+static void
+gtp_current_set_check(void)
+{
+ if (current_gtp_src == NULL)
+ current_gtp = current_gtp->next;
+}
+
+static void
+gtp_current_action_check(void)
+{
+ if (current_gtp_action == NULL) {
+ current_gtp_src = current_gtp->src;
+ gtp_current_set_check();
+ }
+}
+
+static int
+gtp_gdbrsp_qtfp(void)
+{
+ if (gtp_list) {
+ current_gtp = gtp_list;
+ gtp_report_tracepoint(current_gtp, gtp_rw_bufp);
+ gtp_rw_size += strlen(gtp_rw_bufp);
+ gtp_rw_bufp += strlen(gtp_rw_bufp);
+ current_gtp_action = current_gtp->action_list;
+ gtp_current_action_check();
+ } else {
+ gtp_rw_bufp[0] = 'l';
+ gtp_rw_size += 1;
+ gtp_rw_bufp += 1;
+ }
+
+ return 1;
+}
+
+static int
+gtp_gdbrsp_qtsp(void)
+{
+ if (current_gtp_action) {
+ gtp_report_action(current_gtp, current_gtp_action,
+ gtp_rw_bufp);
+ gtp_rw_size += strlen(gtp_rw_bufp);
+ gtp_rw_bufp += strlen(gtp_rw_bufp);
+ current_gtp_action = current_gtp_action->next;
+ gtp_current_action_check();
+ goto out;
+ }
+
+ if (current_gtp_src) {
+ gtp_report_src(current_gtp, current_gtp_src, gtp_rw_bufp);
+ gtp_rw_size += strlen(gtp_rw_bufp);
+ gtp_rw_bufp += strlen(gtp_rw_bufp);
+ current_gtp_src = current_gtp_src->next;
+ gtp_current_set_check();
+ goto out;
+ }
+
+ if (current_gtp) {
+ gtp_report_tracepoint(current_gtp, gtp_rw_bufp);
+ gtp_rw_size += strlen(gtp_rw_bufp);
+ gtp_rw_bufp += strlen(gtp_rw_bufp);
+ current_gtp_action = current_gtp->action_list;
+ gtp_current_action_check();
+ } else {
+ gtp_rw_bufp[0] = 'l';
+ gtp_rw_size += 1;
+ gtp_rw_bufp += 1;
+ }
+out:
+ return 1;
+}
+
+static void
+gtp_report_var(void)
+{
+ sprintf(gtp_rw_bufp, "%x:%s", current_gtp_var->num,
+ current_gtp_var->src);
+ gtp_rw_size += strlen(gtp_rw_bufp);
+ gtp_rw_bufp += strlen(gtp_rw_bufp);
+}
+
+static int
+gtp_gdbrsp_qtfsv(int f)
+{
+ if (f)
+ current_gtp_var = gtp_var_list;
+
+ if (current_gtp_var) {
+ gtp_report_var();
+ current_gtp_var = current_gtp_var->next;
+ } else {
+ gtp_rw_bufp[0] = 'l';
+ gtp_rw_size += 1;
+ gtp_rw_bufp += 1;
+ }
+
+ return 1;
+}
+
+static int
+gtp_gdbrsp_qtv(char *pkg)
+{
+ ULONGEST num;
+ struct gtp_var *var = NULL;
+ struct gtp_frame_var *vr = NULL;
+ uint64_t val = 0;
+
+ pkg = hex2ulongest(pkg, &num);
+
+ if (num == GTP_VAR_CPU_NUMBER_ID) {
+ val = (uint64_t)gtp_cpu_number;
+ goto output_value;
+ } else if (num == GTP_VAR_LAST_ERRNO_ID) {
+ val = (uint64_t)gtp_start_last_errno;
+ goto output_value;
+ }
+
+#ifdef GTP_FRAME_SIMPLE
+ if (gtp_start || !gtp_frame_current) {
+#else
+ if (gtp_start || gtp_frame_current_num < 0) {
+#endif
+ if (num == GTP_VAR_CLOCK_ID) {
+ val = (uint64_t)GTP_LOCAL_CLOCK;
+ goto output_value;
+ } else if (num == GTP_VAR_RDTSC_ID) {
+ unsigned long long a;
+ rdtscll(a);
+ val = (uint64_t)a;
+ goto output_value;
+ } else if (num == GTP_VAR_XTIME_SEC_ID
+ || num == GTP_VAR_XTIME_NSEC_ID) {
+ struct timespec time;
+
+ getnstimeofday(&time);
+ if (num == GTP_VAR_XTIME_SEC_ID)
+ val = (uint64_t)time.tv_sec;
+ else
+ val = (uint64_t)time.tv_nsec;
+
+ goto output_value;
+ }
+
+ if (GTP_VAR_IS_SPECIAL(num))
+ goto out;
+ var = gtp_var_find(num);
+ if (var == NULL)
+ goto out;
+#ifdef CONFIG_PERF_EVENTS
+ if (var->ptid == pe_tv_val
+ || var->ptid == pe_tv_enabled
+ || var->ptid == pe_tv_running) {
+ if (gtp_start)
+ var->pts->val =
+ perf_event_read_value(var->pts->event,
+ &(var->pts->enabled),
+ &(var->pts->running));
+ switch (var->ptid) {
+ case pe_tv_val:
+ val = (uint64_t)(var->pts->val);
+ break;
+ case pe_tv_enabled:
+ val = (uint64_t)(var->pts->enabled);
+ break;
+ case pe_tv_running:
+ val = (uint64_t)(var->pts->running);
+ break;
+ default:
+ break;
+ }
+ goto out;
+ }
+#endif
+ val = var->val;
+ } else {
+#ifdef GTP_FRAME_SIMPLE
+ char *next;
+
+ for (next = *(char **)(gtp_frame_current + FID_SIZE); next;
+ next = *(char **)(next + FID_SIZE)) {
+ if (FID(next) == FID_VAR) {
+ vr = (struct gtp_frame_var *)
+ (next + FID_SIZE + sizeof(char *));
+ if (vr->num == (unsigned int)num)
+ goto while_stop;
+ }
+ }
+#else
+ int is_first = 1;
+ struct ring_buffer_event *rbe;
+ char *tmp;
+
+ /* Handle $cpu_id and $clock. */
+ if (GTP_VAR_AUTO_TRACEV(num)) {
+ if (num == GTP_VAR_CLOCK_ID)
+ val = gtp_frame_current_clock;
+ else if (num == GTP_VAR_CPU_ID)
+ val = gtp_frame_current_cpu;
+ goto output_value;
+ }
+re_find:
+ while (1) {
+ rbe = ring_buffer_iter_peek
+ (gtp_frame_iter[gtp_frame_current_cpu], NULL);
+ if (rbe == NULL)
+ break;
+ tmp = ring_buffer_event_data(rbe);
+ if (FID(tmp) == FID_HEAD)
+ break;
+ if (FID(tmp) == FID_VAR) {
+ vr = (struct gtp_frame_var *)(tmp + FID_SIZE);
+ if (vr->num == (unsigned int)num)
+ goto while_stop;
+ }
+ ring_buffer_read(gtp_frame_iter[gtp_frame_current_cpu],
+ NULL);
+ }
+ if (is_first) {
+ gtp_frame_head_find_num(gtp_frame_current_num);
+ ring_buffer_read(gtp_frame_iter[gtp_frame_current_cpu],
+ NULL);
+ is_first = 0;
+ goto re_find;
+ }
+#endif
+ vr = NULL;
+while_stop:
+ if (vr)
+ val = vr->val;
+ }
+
+out:
+ if (var || vr) {
+output_value:
+ sprintf(gtp_rw_bufp, "V%08x%08x",
+ (unsigned int) (val >> 32),
+ (unsigned int) (val & 0xffffffff));
+ gtp_rw_size += strlen(gtp_rw_bufp);
+ gtp_rw_bufp += strlen(gtp_rw_bufp);
+ } else {
+ gtp_rw_bufp[0] = 'U';
+ gtp_rw_size += 1;
+ gtp_rw_bufp += 1;
+ }
+
+ return 1;
+}
+
+static int
+gtp_gdbrsp_qT(char *pkg)
+{
+ int ret = 1;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_qT: %s\n", pkg);
+#endif
+
+ if (strcmp("Status", pkg) == 0)
+ ret = gtp_gdbrsp_qtstatus();
+ else if (strcmp("fP", pkg) == 0)
+ ret = gtp_gdbrsp_qtfp();
+ else if (strcmp("sP", pkg) == 0)
+ ret = gtp_gdbrsp_qtsp();
+ else if (strcmp("fV", pkg) == 0)
+ ret = gtp_gdbrsp_qtfsv(1);
+ else if (strcmp("sV", pkg) == 0)
+ ret = gtp_gdbrsp_qtfsv(0);
+ else if (strncmp("V:", pkg, 2) == 0)
+ ret = gtp_gdbrsp_qtv(pkg + 2);
+
+ return ret;
+}
+
+static uint8_t gtp_m_buffer[0xffff];
+
+static int
+gtp_gdbrsp_m(char *pkg)
+{
+ int i;
+ ULONGEST addr, len;
+
+ /* Get add and len. */
+ if (pkg[0] == '\0')
+ return -EINVAL;
+ pkg = hex2ulongest(pkg, &addr);
+ if (pkg[0] != ',')
+ return -EINVAL;
+ pkg++;
+ pkg = hex2ulongest(pkg, &len);
+ if (len == 0)
+ return -EINVAL;
+ len &= 0xffff;
+ len = (ULONGEST) min((int)((GTP_RW_MAX - 4 - gtp_rw_size) / 2),
+ (int)len);
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_m: addr = 0x%lx len = %d\n",
+ (unsigned long) addr, (int) len);
+#endif
+
+#ifdef GTP_FRAME_SIMPLE
+ if (gtp_start || !gtp_frame_current) {
+#else
+ if (gtp_start || gtp_frame_current_num < 0) {
+#endif
+ if (probe_kernel_read(gtp_m_buffer, (void *)(CORE_ADDR)addr,
+ (size_t)len))
+ return -EFAULT;
+ } else {
+#ifdef GTP_FRAME_SIMPLE
+ char *next;
+#endif
+ int ret;
+
+ /* The following part is for gtpro support.
+ It is not available because it make disassemble cannot
+ work when select a trace frame. */
+#if 0
+ struct gtpro_entry *gtroe;
+
+ memset(gtp_m_buffer, 0, len);
+
+ /* Read the gtpro. */
+ for (gtroe = gtpro_list; gtroe; gtroe = gtroe->next) {
+ CORE_ADDR cur_start, cur_end;
+
+ cur_start = max(gtroe->start, (CORE_ADDR)addr);
+ cur_end = min(gtroe->end, ((CORE_ADDR)(addr + len)));
+ if (cur_start < cur_end) {
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_m: ro read "
+ "start = 0x%lx end = 0x%lx\n",
+ (unsigned long) cur_start,
+ (unsigned long) cur_end);
+#endif
+ if (probe_kernel_read(gtp_m_buffer,
+ (void *)cur_start,
+ (size_t)(cur_end
+ - cur_start)))
+ return -EFAULT;
+ }
+ }
+#endif
+ ret = probe_kernel_read(gtp_m_buffer, (void *)(CORE_ADDR)addr,
+ (size_t)len);
+#ifdef GTP_FRAME_SIMPLE
+ for (next = *(char **)(gtp_frame_current + FID_SIZE); next;
+ next = *(char **)(next + FID_SIZE)) {
+ if (FID(next) == FID_MEM) {
+ struct gtp_frame_mem *mr;
+ ULONGEST cur_start, cur_end;
+ uint8_t *buf;
+
+ mr = (struct gtp_frame_mem *)
+ (next + FID_SIZE + sizeof(char *));
+ buf = next + GTP_FRAME_MEM_SIZE;
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_m: section "
+ "addr = 0x%lx size = %lu\n",
+ (unsigned long) mr->addr,
+ (unsigned long) mr->size);
+#endif
+ cur_start = max(((ULONGEST)mr->addr), addr);
+ cur_end = min(((ULONGEST)mr->addr
+ + mr->size),
+ (addr + len));
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_m: read "
+ "start = 0x%lx end = 0x%lx\n",
+ (unsigned long) cur_start,
+ (unsigned long) cur_end);
+#endif
+ if (cur_start < cur_end) {
+ memcpy(gtp_m_buffer,
+ buf + cur_start - mr->addr,
+ cur_end - cur_start);
+ ret = 0;
+ }
+ }
+ }
+#else
+ gtp_frame_head_find_num(gtp_frame_current_num);
+ ring_buffer_read(gtp_frame_iter[gtp_frame_current_cpu], NULL);
+
+ while (1) {
+ struct ring_buffer_event *rbe;
+ char *tmp;
+
+ rbe = ring_buffer_iter_peek
+ (gtp_frame_iter[gtp_frame_current_cpu], NULL);
+ if (rbe == NULL)
+ break;
+ tmp = ring_buffer_event_data(rbe);
+ if (FID(tmp) == FID_HEAD)
+ break;
+ if (FID(tmp) == FID_MEM) {
+ struct gtp_frame_mem *mr;
+ ULONGEST cur_start, cur_end;
+ uint8_t *buf;
+
+ mr = (struct gtp_frame_mem *)
+ (tmp + FID_SIZE);
+ buf = tmp + GTP_FRAME_MEM_SIZE;
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_m: section "
+ "addr = 0x%lx size = %lu\n",
+ (unsigned long) mr->addr,
+ (unsigned long) mr->size);
+#endif
+ cur_start = max(((ULONGEST)mr->addr), addr);
+ cur_end = min(((ULONGEST)mr->addr
+ + mr->size),
+ (addr + len));
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_m: read "
+ "start = 0x%lx end = 0x%lx\n",
+ (unsigned long) cur_start,
+ (unsigned long) cur_end);
+#endif
+ if (cur_start < cur_end) {
+ memcpy(gtp_m_buffer,
+ buf + cur_start - mr->addr,
+ cur_end - cur_start);
+ ret = 0;
+ }
+ }
+ ring_buffer_read(gtp_frame_iter[gtp_frame_current_cpu],
+ NULL);
+ }
+#endif
+ if (ret)
+ return -EFAULT;
+ }
+
+ for (i = 0; i < (int)len; i++) {
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_gdbrsp_m: %d %02x\n", i, gtp_m_buffer[i]);
+#endif
+ sprintf(gtp_rw_bufp, "%02x", gtp_m_buffer[i]);
+ gtp_rw_bufp += 2;
+ gtp_rw_size += 2;
+ }
+
+ return 1;
+}
+
+static int
+gtp_gdbrsp_g(void)
+{
+#ifdef GTP_FRAME_SIMPLE
+ char *next;
+#endif
+ struct pt_regs *regs;
+
+ if (GTP_RW_MAX - 4 - gtp_rw_size < GTP_REG_ASCII_SIZE)
+ return -E2BIG;
+
+#ifdef GTP_FRAME_SIMPLE
+ if (gtp_start || !gtp_frame_current) {
+#else
+ if (gtp_start || gtp_frame_current_num < 0) {
+#endif
+ memset(gtp_rw_bufp, '0', GTP_REG_ASCII_SIZE);
+ goto out;
+ }
+
+ /* Get the regs. */
+ regs = NULL;
+#ifdef GTP_FRAME_SIMPLE
+ for (next = *(char **)(gtp_frame_current + FID_SIZE); next;
+ next = *(char **)(next + FID_SIZE)) {
+ if (FID(next) == FID_REG) {
+ regs = (struct pt_regs *)
+ (next + FID_SIZE + sizeof(char *));
+ break;
+ }
+ }
+#else
+ {
+ int is_first = 1;
+ struct ring_buffer_event *rbe;
+ char *tmp;
+
+re_find:
+ while (1) {
+ rbe = ring_buffer_iter_peek
+ (gtp_frame_iter[gtp_frame_current_cpu], NULL);
+ if (rbe == NULL)
+ break;
+ tmp = ring_buffer_event_data(rbe);
+ if (FID(tmp) == FID_HEAD)
+ break;
+ if (FID(tmp) == FID_REG) {
+ regs = (struct pt_regs *)(tmp + FID_SIZE);
+ is_first = 0;
+ break;
+ }
+ ring_buffer_read(gtp_frame_iter[gtp_frame_current_cpu],
+ NULL);
+ }
+ if (is_first) {
+ gtp_frame_head_find_num(gtp_frame_current_num);
+ ring_buffer_read(gtp_frame_iter[gtp_frame_current_cpu],
+ NULL);
+ is_first = 0;
+ goto re_find;
+ }
+ }
+#endif
+ if (regs)
+ gtp_regs2ascii(regs, gtp_rw_bufp);
+ else {
+ struct pt_regs pregs;
+ struct gtp_entry *tpe;
+
+ memset(&pregs, '\0', sizeof(struct pt_regs));
+ tpe = gtp_list_find_without_addr(gtp_frame_current_tpe);
+ if (tpe)
+ GTP_REGS_PC(&pregs) = (unsigned long)tpe->addr;
+ gtp_regs2ascii(&pregs, gtp_rw_bufp);
+ }
+out:
+ gtp_rw_bufp += GTP_REG_ASCII_SIZE;
+ gtp_rw_size += GTP_REG_ASCII_SIZE;
+
+ return 1;
+}
+
+static DEFINE_SEMAPHORE(gtp_rw_lock);
+static DECLARE_WAIT_QUEUE_HEAD(gtp_rw_wq);
+static unsigned int gtp_rw_count;
+static unsigned int gtp_frame_count;
+
+static void
+gtp_frame_count_release(void)
+{
+ gtp_frame_count--;
+ if (gtp_frame_count == 0) {
+ if (!gtp_disconnected_tracing) {
+ gtp_gdbrsp_qtstop();
+ gtp_gdbrsp_qtinit();
+ if (gtp_frame) {
+#ifdef GTP_FRAME_SIMPLE
+ vfree(gtp_frame);
+#else
+ ring_buffer_free(gtp_frame);
+#endif
+ gtp_frame = NULL;
+ }
+ }
+ }
+}
+
+static int
+gtp_open(struct inode *inode, struct file *file)
+{
+ int ret = 0;
+
+ down(>p_rw_lock);
+ if (gtp_gtp_pid >= 0) {
+ if (get_current()->pid != gtp_gtp_pid) {
+ ret = -EBUSY;
+ goto out;
+ }
+ }
+
+ if (gtp_rw_count == 0) {
+ gtp_read_ack = 0;
+ gtp_rw_buf = vmalloc(GTP_RW_MAX);
+ if (!gtp_rw_buf) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ }
+ gtp_rw_count++;
+
+ gtp_frame_count++;
+
+ gtp_gtp_pid_count++;
+ if (gtp_gtp_pid < 0)
+ gtp_gtp_pid = get_current()->pid;
+
+out:
+ up(>p_rw_lock);
+ return ret;
+}
+
+static int
+gtp_release(struct inode *inode, struct file *file)
+{
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_release\n");
+#endif
+
+ down(>p_rw_lock);
+ gtp_rw_count--;
+ if (gtp_rw_count == 0)
+ vfree(gtp_rw_buf);
+
+ gtp_frame_count_release();
+
+ gtp_gtp_pid_count--;
+ if (gtp_gtp_pid_count == 0)
+ gtp_gtp_pid = -1;
+
+ up(>p_rw_lock);
+
+ return 0;
+}
+
+static long
+gtp_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
+{
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_ioctl: %x\n", cmd);
+#endif
+
+ return 0;
+}
+
+static ssize_t
+gtp_write(struct file *file, const char __user *buf, size_t size,
+ loff_t *ppos)
+{
+ char *rsppkg = NULL;
+ int i, ret;
+ unsigned char csum = 0;
+
+ if (down_interruptible(>p_rw_lock))
+ return -EINTR;
+
+ if (size == 0) {
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_write: try write 0 size.\n");
+#endif
+ goto error_out;
+ }
+
+ size = min_t(size_t, size, GTP_RW_MAX);
+ if (copy_from_user(gtp_rw_buf, buf, size)) {
+ size = -EFAULT;
+ goto error_out;
+ }
+
+ if (gtp_rw_buf[0] == '+' || gtp_rw_buf[0] == '-'
+ || gtp_rw_buf[0] == '\3') {
+ if (gtp_rw_buf[0] == '+')
+ gtp_rw_size = 0;
+ size = 1;
+ goto out;
+ }
+
+ if (size < 4) {
+ gtp_read_ack = '-';
+ goto out;
+ }
+ /* Check format and crc and get the rsppkg. */
+ for (i = 0; i < size - 2; i++) {
+ if (rsppkg == NULL) {
+ if (gtp_rw_buf[i] == '$')
+ rsppkg = gtp_rw_buf + i + 1;
+ } else {
+ if (gtp_rw_buf[i] == '#')
+ break;
+ else
+ csum += gtp_rw_buf[i];
+ }
+ }
+ if (rsppkg && gtp_rw_buf[i] == '#') {
+ /* Format is OK. Check crc. */
+ unsigned char c1, c2;
+
+ gtp_rw_buf[i] = '\0';
+
+ c1 = gtp_rw_buf[i+1];
+ c2 = gtp_rw_buf[i+2];
+ if (csum == (c1 << 4) + c2) {
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_write: crc error\n");
+#endif
+ gtp_read_ack = '-';
+ goto out;
+ }
+ } else {
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_write: format error\n");
+#endif
+ gtp_read_ack = '-';
+ goto out;
+ }
+ gtp_read_ack = '+';
+ size = i + 3;
+
+ wake_up_interruptible_nr(>p_rw_wq, 1);
+
+ up(>p_rw_lock);
+ if (down_interruptible(>p_rw_lock))
+ return -EINTR;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_write: %s\n", rsppkg);
+#endif
+
+ /* Handle rsppkg and put return to gtp_rw_buf. */
+ gtp_rw_buf[0] = '$';
+ gtp_rw_bufp = gtp_rw_buf + 1;
+ gtp_rw_size = 0;
+ ret = 1;
+ switch (rsppkg[0]) {
+ case '?':
+ strcpy(gtp_rw_bufp, "S05");
+ gtp_rw_bufp += 3;
+ gtp_rw_size += 3;
+ break;
+ case 'g':
+ ret = gtp_gdbrsp_g();
+ break;
+ case 'm':
+ ret = gtp_gdbrsp_m(rsppkg + 1);
+ break;
+ case 'Q':
+ if (rsppkg[1] == 'T')
+ ret = gtp_gdbrsp_QT(rsppkg + 2);
+ break;
+ case 'q':
+ if (rsppkg[1] == 'T')
+ ret = gtp_gdbrsp_qT(rsppkg + 2);
+ else if (strncmp("qSupported", rsppkg, 10) == 0) {
+ strcpy(gtp_rw_bufp,
+ "ConditionalTracepoints+;"
+ "TracepointSource+;DisconnectedTracing+");
+ gtp_rw_size += strlen(gtp_rw_bufp);
+ gtp_rw_bufp += strlen(gtp_rw_bufp);
+ ret = 1;
+ }
+ break;
+ case 's':
+ case 'S':
+ case 'c':
+ case 'C':
+ ret = -1;
+ break;
+ }
+ if (ret == 0) {
+ strcpy(gtp_rw_bufp, "OK");
+ gtp_rw_bufp += 2;
+ gtp_rw_size += 2;
+ } else if (ret < 0) {
+ sprintf(gtp_rw_bufp, "E%02x", -ret);
+ gtp_rw_bufp += 3;
+ gtp_rw_size += 3;
+ }
+
+ gtp_rw_bufp[0] = '#';
+ csum = 0;
+ for (i = 1; i < gtp_rw_size + 1; i++)
+ csum += gtp_rw_buf[i];
+ gtp_rw_bufp[1] = TOHEX(csum >> 4);
+ gtp_rw_bufp[2] = TOHEX(csum & 0x0f);
+ gtp_rw_bufp = gtp_rw_buf;
+ gtp_rw_size += 4;
+
+out:
+ wake_up_interruptible_nr(>p_rw_wq, 1);
+error_out:
+ up(>p_rw_lock);
+ return size;
+}
+
+static ssize_t
+gtp_read(struct file *file, char __user *buf, size_t size,
+ loff_t *ppos)
+{
+ int err;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_read\n");
+#endif
+
+ if (size == 0)
+ goto out;
+
+ if (down_interruptible(>p_rw_lock))
+ return -EINTR;
+
+ if (gtp_read_ack) {
+ err = put_user(gtp_read_ack, buf);
+ if (err) {
+ size = -err;
+ goto out;
+ }
+ gtp_read_ack = 0;
+ size = 1;
+ goto out;
+ }
+
+ size = min(gtp_rw_size, size);
+ if (size == 0)
+ goto out;
+ if (copy_to_user(buf, gtp_rw_bufp, size)) {
+ size = -EFAULT;
+ goto out;
+ }
+ gtp_rw_bufp += size;
+ gtp_rw_size -= size;
+
+out:
+ up(>p_rw_lock);
+ return size;
+}
+
+static unsigned int
+gtp_poll(struct file *file, poll_table *wait)
+{
+ unsigned int mask = POLLOUT | POLLWRNORM;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtp_poll\n");
+#endif
+
+ down(>p_rw_lock);
+ poll_wait(file, >p_rw_wq, wait);
+ if (gtp_read_ack || gtp_rw_size)
+ mask |= POLLIN | POLLRDNORM;
+ up(>p_rw_lock);
+
+ return mask;
+}
+
+struct gtp_realloc_s {
+ char *buf;
+ size_t size;
+ size_t real_size;
+};
+
+static char *
+gtp_realloc(struct gtp_realloc_s *grs, size_t size, int is_end)
+{
+ char *tmp;
+
+ if ((grs->real_size < grs->size + size)
+ || (is_end && grs->real_size != grs->size + size)) {
+ grs->real_size = grs->size + size;
+ if (!is_end)
+ grs->real_size += 100;
+
+ tmp = vmalloc(grs->real_size);
+ if (!tmp) {
+ vfree(grs->buf);
+ memset(grs, 0, sizeof(struct gtp_realloc_s));
+ return NULL;
+ }
+
+ memcpy(tmp, grs->buf, grs->size);
+ vfree(grs->buf);
+ grs->buf = tmp;
+ }
+
+ return grs->buf + grs->size;
+}
+
+static int
+gtp_frame2file_r(struct gtp_realloc_s *grs, uint32_t *data_size, char *frame)
+{
+ char *wbuf;
+
+ wbuf = gtp_realloc(grs, GTP_REG_BIN_SIZE + 1, 0);
+ if (!wbuf)
+ return -1;
+
+ wbuf[0] = 'R';
+#ifdef GTP_FRAME_SIMPLE
+ gtp_regs2bin((struct pt_regs *)(frame + FID_SIZE + sizeof(char *)),
+ wbuf + 1);
+#else
+ gtp_regs2bin((struct pt_regs *)(frame + FID_SIZE), wbuf + 1);
+#endif
+ grs->size += GTP_REG_BIN_SIZE + 1;
+
+ *data_size += GTP_REG_BIN_SIZE + 1;
+
+ return 0;
+}
+
+static int
+gtp_frame2file_m(struct gtp_realloc_s *grs, uint32_t *data_size, char *frame)
+{
+ struct gtp_frame_mem *mr;
+ uint8_t *buf;
+ ULONGEST addr;
+ size_t remaining;
+
+#ifdef GTP_FRAME_SIMPLE
+ mr = (struct gtp_frame_mem *) (frame + FID_SIZE + sizeof(char *));
+#else
+ mr = (struct gtp_frame_mem *) (frame + FID_SIZE);
+#endif
+ buf = frame + GTP_FRAME_MEM_SIZE;
+ addr = mr->addr;
+ remaining = mr->size;
+
+ while (remaining > 0) {
+ uint16_t blocklen;
+ char *wbuf;
+ size_t sp;
+
+ blocklen = remaining > 65535 ? 65535 : remaining;
+
+ sp = 1 + sizeof(addr) + sizeof(blocklen) + blocklen;
+ wbuf = gtp_realloc(grs, sp, 0);
+ if (!wbuf)
+ return -1;
+
+ wbuf[0] = 'M';
+ wbuf += 1;
+
+ memcpy(wbuf, &addr, sizeof(addr));
+ wbuf += sizeof(addr);
+
+ memcpy(wbuf, &blocklen, sizeof(blocklen));
+ wbuf += sizeof(blocklen);
+
+ memcpy(wbuf, buf, blocklen);
+
+ addr += blocklen;
+ remaining -= blocklen;
+ buf += blocklen;
+
+ grs->size += sp;
+ *data_size += sp;
+ }
+
+ return 0;
+}
+
+static int
+gtp_frame2file_v(struct gtp_realloc_s *grs, uint32_t *data_size, char *frame)
+{
+ struct gtp_frame_var *vr;
+ size_t sp = 1 + sizeof(unsigned int)
+ + sizeof(uint64_t);
+ char *wbuf;
+
+ wbuf = gtp_realloc(grs, sp, 0);
+ if (!wbuf)
+ return -1;
+
+#ifdef GTP_FRAME_SIMPLE
+ vr = (struct gtp_frame_var *) (frame + FID_SIZE + sizeof(char *));
+#else
+ vr = (struct gtp_frame_var *) (frame + FID_SIZE);
+#endif
+
+ wbuf[0] = 'V';
+ wbuf += 1;
+
+ memcpy(wbuf, &vr->num, sizeof(unsigned int));
+ wbuf += sizeof(unsigned int);
+
+ memcpy(wbuf, &vr->val, sizeof(uint64_t));
+ wbuf += sizeof(uint64_t);
+
+ grs->size += sp;
+ *data_size += sp;
+
+ return 0;
+}
+
+static int
+#ifdef GTP_FRAME_SIMPLE
+gtp_frame2file(struct gtp_realloc_s *grs, char *frame)
+#else
+gtp_frame2file(struct gtp_realloc_s *grs, int cpu)
+#endif
+{
+ int16_t *tmp16p;
+ char *next;
+ char *wbuf;
+ uint32_t data_size;
+#ifndef GTP_FRAME_SIMPLE
+ struct ring_buffer_event *rbe;
+ u64 clock;
+#endif
+
+ /* Head. */
+ tmp16p = (int16_t *)gtp_realloc(grs, 2, 0);
+ if (!tmp16p)
+ return -1;
+#ifdef GTP_FRAME_SIMPLE
+ *tmp16p = (int16_t)*(ULONGEST *)(frame + FID_SIZE + sizeof(char *));
+#else
+ rbe = ring_buffer_read(gtp_frame_iter[cpu], &clock);
+ if (rbe == NULL) {
+ /* It will not happen, just for safe. */
+ return -1;
+ }
+ next = ring_buffer_event_data(rbe);
+ *tmp16p = (int16_t)*(ULONGEST *)(next + FID_SIZE);
+#endif
+ grs->size += 2;
+ /* This part is for the data_size. */
+ wbuf = gtp_realloc(grs, 4, 0);
+ if (!wbuf)
+ return -1;
+ grs->size += 4;
+
+ /* Body. */
+ data_size = 0;
+
+#ifndef GTP_FRAME_SIMPLE
+ {
+ /* Handle $cpu_id and $clock. */
+ struct gtp_frame_var *vr;
+ char frame[GTP_FRAME_VAR_SIZE];
+
+
+ vr = (struct gtp_frame_var *) (frame + FID_SIZE);
+ vr->num = GTP_VAR_CLOCK_ID;
+ vr->val = clock;
+ if (gtp_frame2file_v(grs, &data_size, frame))
+ return -1;
+ vr->num = GTP_VAR_CPU_ID;
+ vr->val = cpu;
+ if (gtp_frame2file_v(grs, &data_size, frame))
+ return -1;
+ }
+#endif
+
+#ifdef GTP_FRAME_SIMPLE
+ for (next = *(char **)(frame + FID_SIZE); next;
+ next = *(char **)(next + FID_SIZE)) {
+#else
+ while (1) {
+ rbe = ring_buffer_iter_peek(gtp_frame_iter[cpu], NULL);
+ if (rbe == NULL)
+ break;
+ next = ring_buffer_event_data(rbe);
+#endif
+ switch (FID(next)) {
+ case FID_REG:
+ if (gtp_frame2file_r(grs, &data_size, next))
+ return -1;
+ break;
+
+ case FID_MEM:
+ if (gtp_frame2file_m(grs, &data_size, next))
+ return -1;
+ break;
+
+ case FID_VAR:
+ if (gtp_frame2file_v(grs, &data_size, next))
+ return -1;
+ break;
+
+#ifndef GTP_FRAME_SIMPLE
+ case FID_HEAD:
+ goto out;
+ break;
+#endif
+ }
+#ifndef GTP_FRAME_SIMPLE
+ ring_buffer_read(gtp_frame_iter[cpu], NULL);
+#endif
+ }
+
+#ifndef GTP_FRAME_SIMPLE
+out:
+#endif
+ /* Set the data_size. */
+ memcpy(grs->buf + grs->size - data_size - 4,
+ &data_size, 4);
+
+ return 0;
+}
+
+static int
+gtp_frame_file_header(struct gtp_realloc_s *grs, int is_end)
+{
+ char *wbuf;
+ struct gtp_entry *tpe;
+ struct gtp_var *tvar;
+ int tmpsize;
+ char tmpbuf[200];
+ int ret = -1;
+
+ /* Head. */
+ wbuf = gtp_realloc(grs, 8, 0);
+ strcpy(wbuf, "\x7fTRACE0\n");
+ grs->size += 8;
+
+ /* BUG: will be a new value. */
+ snprintf(tmpbuf, 200, "R %x\n", GTP_REG_BIN_SIZE);
+ wbuf = gtp_realloc(grs, strlen(tmpbuf), 0);
+ if (!wbuf)
+ goto out;
+ strcpy(wbuf, tmpbuf);
+ grs->size += strlen(tmpbuf);
+
+ strcpy(tmpbuf, "status 0;");
+ wbuf = gtp_realloc(grs, strlen(tmpbuf), 0);
+ if (!wbuf)
+ goto out;
+ strcpy(wbuf, tmpbuf);
+ grs->size += strlen(tmpbuf);
+
+ for (tpe = gtp_list; tpe; tpe = tpe->next) {
+ if (tpe->reason != gtp_stop_normal)
+ break;
+ }
+ tmpsize = gtp_get_status(tpe, tmpbuf);
+ wbuf = gtp_realloc(grs, tmpsize, 0);
+ if (!wbuf)
+ goto out;
+ memcpy(wbuf, tmpbuf, tmpsize);
+ grs->size += tmpsize;
+
+ wbuf = gtp_realloc(grs, 1, 0);
+ if (!wbuf)
+ goto out;
+ wbuf[0] = '\n';
+ grs->size += 1;
+
+ /* Tval. */
+ for (tvar = gtp_var_list; tvar; tvar = tvar->next) {
+ snprintf(tmpbuf, 200, "tsv %x:%s\n", tvar->num, tvar->src);
+ wbuf = gtp_realloc(grs, strlen(tmpbuf), 0);
+ if (!wbuf)
+ goto out;
+ strcpy(wbuf, tmpbuf);
+ grs->size += strlen(tmpbuf);
+ }
+
+ /* Tracepoint. */
+ for (tpe = gtp_list; tpe; tpe = tpe->next) {
+ struct action *ae;
+ struct gtpsrc *src;
+
+ /* Tpe. */
+ gtp_report_tracepoint(tpe, tmpbuf);
+ wbuf = gtp_realloc(grs, strlen(tmpbuf) + 5, 0);
+ if (!wbuf)
+ goto out;
+ sprintf(wbuf, "tp %s\n", tmpbuf);
+ grs->size += strlen(tmpbuf) + 4;
+ /* Action. */
+ for (ae = tpe->action_list; ae; ae = ae->next) {
+ gtp_report_action(tpe, ae, tmpbuf);
+ wbuf = gtp_realloc(grs, strlen(tmpbuf) + 5, 0);
+ if (!wbuf)
+ goto out;
+ sprintf(wbuf, "tp %s\n", tmpbuf);
+ grs->size += strlen(tmpbuf) + 4;
+ }
+ /* Src. */
+ for (src = tpe->src; src; src = src->next) {
+ gtp_report_src(tpe, src, tmpbuf);
+ wbuf = gtp_realloc(grs, strlen(tmpbuf) + 5, 0);
+ if (!wbuf)
+ goto out;
+ sprintf(wbuf, "tp %s\n", tmpbuf);
+ grs->size += strlen(tmpbuf) + 4;
+ }
+ }
+
+ wbuf = gtp_realloc(grs, 1, is_end);
+ if (!wbuf)
+ goto out;
+ wbuf[0] = '\n';
+ grs->size += 1;
+
+ ret = 0;
+out:
+ return ret;
+}
+
+static ssize_t
+gtpframe_read(struct file *file, char __user *buf, size_t size,
+ loff_t *ppos)
+{
+ ssize_t ret = -ENOMEM;
+#ifndef GTP_FRAME_SIMPLE
+ int old_num = -2;
+#endif
+
+recheck:
+ down(>p_rw_lock);
+ if (gtp_start) {
+ up(>p_rw_lock);
+ if (wait_event_interruptible(gtpframe_wq,
+ !gtp_start) == -ERESTARTSYS)
+ return -ERESTARTSYS;
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtpframe_read: goto recheck\n");
+#endif
+ goto recheck;
+ }
+
+ /* Set gtp_frame_file if need. */
+ if (!gtp_frame_file) {
+ char *wbuf;
+#ifdef GTP_FRAME_SIMPLE
+ char *frame;
+#endif
+ struct gtp_realloc_s gr;
+
+#ifdef GTP_FRAME_SIMPLE
+ if (gtp_frame_is_circular)
+ gr.real_size = GTP_FRAME_SIZE;
+ else
+ gr.real_size = gtp_frame_w_start - gtp_frame;
+#else
+ gr.real_size =
+ ring_buffer_entries(gtp_frame) * GTP_FRAME_HEAD_SIZE;
+#endif
+ gr.real_size += 200;
+ gr.buf = vmalloc(gr.real_size);
+ if (!gr.buf)
+ goto out;
+ gr.size = 0;
+
+ if (gtp_frame_file_header(&gr, 0))
+ goto out;
+
+ /* Frame. */
+ if (atomic_read(>p_frame_create) == 0)
+ goto end;
+#ifdef GTP_FRAME_SIMPLE
+ frame = gtp_frame_r_start;
+ do {
+ if (frame == gtp_frame_end)
+ frame = gtp_frame;
+
+ if (FID(frame) == FID_HEAD) {
+ if (gtp_frame2file(&gr, frame))
+ goto out;
+ }
+
+ frame = gtp_frame_next(frame);
+ if (!frame)
+ break;
+ } while (frame != gtp_frame_w_start);
+#else
+ old_num = gtp_frame_current_num;
+ gtp_frame_iter_reset();
+ while (1) {
+ int cpu;
+
+ cpu = gtp_frame_iter_peek_head();
+ if (cpu < 0)
+ break;
+
+ if (gtp_frame2file(&gr, cpu))
+ goto out;
+ }
+#endif
+
+end:
+ /* End. */
+ wbuf = gtp_realloc(&gr, 2, 1);
+ if (!wbuf)
+ goto out;
+ wbuf[0] = '\0';
+ wbuf[1] = '\0';
+ gr.size += 2;
+
+ gtp_frame_file = gr.buf;
+ gtp_frame_file_size = gr.size;
+ }
+
+ /* Set buf. */
+ ret = size;
+ if (*ppos + ret > gtp_frame_file_size) {
+ ret = gtp_frame_file_size - *ppos;
+ if (ret <= 0) {
+ ret = 0;
+ goto out;
+ }
+ }
+ if (copy_to_user(buf, gtp_frame_file + *ppos, ret)) {
+ size = -EFAULT;
+ goto out;
+ }
+ *ppos += ret;
+
+out:
+#ifndef GTP_FRAME_SIMPLE
+ if (old_num == -1)
+ gtp_frame_iter_reset();
+ else if (old_num >= 0) {
+ gtp_frame_head_find_num(old_num);
+ ring_buffer_read(gtp_frame_iter[gtp_frame_current_cpu], NULL);
+ }
+#endif
+ up(>p_rw_lock);
+ return ret;
+}
+
+static int
+gtpframe_open(struct inode *inode, struct file *file)
+{
+recheck:
+ down(>p_rw_lock);
+ if (!gtp_frame) {
+ up(>p_rw_lock);
+ if (wait_event_interruptible(gtpframe_wq,
+ gtp_frame) == -ERESTARTSYS)
+ return -ERESTARTSYS;
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtpframe_open: goto recheck\n");
+#endif
+ goto recheck;
+ }
+
+ if (gtp_gtpframe_pid >= 0) {
+ if (get_current()->pid != gtp_gtpframe_pid) {
+ up(>p_rw_lock);
+ return -EBUSY;
+ }
+ }
+
+ gtp_frame_count++;
+
+ gtp_gtpframe_pid_count++;
+ if (gtp_gtpframe_pid < 0)
+ gtp_gtpframe_pid = get_current()->pid;
+
+ up(>p_rw_lock);
+ return 0;
+}
+
+static int
+gtpframe_release(struct inode *inode, struct file *file)
+{
+ down(>p_rw_lock);
+ gtp_frame_count_release();
+
+ gtp_gtpframe_pid_count--;
+ if (gtp_gtpframe_pid_count == 0)
+ gtp_gtpframe_pid = -1;
+ up(>p_rw_lock);
+
+ return 0;
+}
+
+#ifndef GTP_FRAME_SIMPLE
+struct gtpframe_pipe_s {
+ loff_t begin;
+ struct gtp_realloc_s *grs;
+ int llseek_move;
+};
+
+static int gtpframe_pipe_is_running;
+
+static int
+gtpframe_pipe_open(struct inode *inode, struct file *file)
+{
+ int ret = -ENOMEM;
+ struct gtpframe_pipe_s *gps = NULL;
+
+ /* Check gtpframe_pipe_is_running. */
+ down(>p_rw_lock);
+ if (gtpframe_pipe_is_running) {
+ up(>p_rw_lock);
+ return -EBUSY;
+ }
+ gtpframe_pipe_is_running = 1;
+
+recheck:
+ if (!gtp_frame) {
+ up(>p_rw_lock);
+ atomic_inc(>pframe_pipe_wq_v);
+ if (wait_event_interruptible(gtpframe_pipe_wq,
+ gtp_frame) == -ERESTARTSYS)
+ return -ERESTARTSYS;
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtpframe_pipe_open: goto recheck\n");
+#endif
+ down(>p_rw_lock);
+ goto recheck;
+ }
+
+ if (gtp_gtpframe_pipe_pid >= 0) {
+ if (get_current()->pid != gtp_gtpframe_pipe_pid) {
+ ret = -EBUSY;
+ goto out;
+ }
+ }
+
+ gps = kmalloc(sizeof(struct gtpframe_pipe_s), GFP_KERNEL);
+ if (gps == NULL)
+ goto out;
+ memset(gps, 0, sizeof(struct gtpframe_pipe_s));
+ gps->grs = kmalloc(sizeof(struct gtp_realloc_s), GFP_KERNEL);
+ if (gps->grs == NULL)
+ goto out;
+ memset(gps->grs, 0, sizeof(struct gtp_realloc_s));
+
+ file->private_data = gps;
+
+ gtp_frame_count++;
+
+ gtp_gtpframe_pipe_pid_count++;
+ if (gtp_gtpframe_pipe_pid < 0)
+ gtp_gtpframe_pipe_pid = get_current()->pid;
+
+ ret = 0;
+out:
+ if (ret && gps) {
+ kfree(gps->grs);
+ kfree(gps);
+ }
+ up(>p_rw_lock);
+ return ret;
+}
+
+static int
+gtpframe_pipe_release(struct inode *inode, struct file *file)
+{
+ struct gtpframe_pipe_s *gps = file->private_data;
+
+ down(>p_rw_lock);
+ gtp_frame_count_release();
+
+ gtp_gtpframe_pipe_pid_count--;
+ if (gtp_gtpframe_pipe_pid_count == 0)
+ gtp_gtpframe_pipe_pid = -1;
+
+ gtpframe_pipe_is_running = 0;
+
+ up(>p_rw_lock);
+
+ if (gps) {
+ if (gps->grs) {
+ if (gps->grs->buf)
+ vfree(gps->grs->buf);
+ kfree(gps->grs);
+ }
+ kfree(gps);
+ }
+
+ return 0;
+}
+
+static int
+gtpframe_pipe_peek(void)
+{
+ u64 min = 0;
+ u64 ts;
+ int cpu;
+ struct ring_buffer_event *rbe;
+ char *next;
+ int ret = -1;
+
+ for_each_online_cpu(cpu) {
+ while (1) {
+ rbe = ring_buffer_peek(gtp_frame, cpu, &ts, NULL);
+ if (rbe == NULL)
+ break;
+ next = ring_buffer_event_data(rbe);
+ if (FID(next) == FID_HEAD)
+ break;
+ ring_buffer_consume(gtp_frame, cpu, &ts, NULL);
+ }
+
+ if (rbe) {
+ if ((min && ts < min) || !min) {
+ min = ts;
+ ret = cpu;
+ }
+ }
+ }
+
+ return ret;
+}
+
+static int
+gtpframe_pipe_get_entry(struct gtp_realloc_s *grs)
+{
+ int cpu;
+ struct ring_buffer_event *rbe;
+ char *next;
+ u64 ts;
+ int16_t *tmp16p;
+ uint32_t data_size;
+
+ grs->size = 0;
+
+recheck:
+#define GTP_PIPE_PEEK (cpu = gtpframe_pipe_peek())
+ GTP_PIPE_PEEK;
+ if (cpu < 0) {
+ /* Didn't get the buffer that have event.
+ Wait and recheck.*/
+ atomic_inc(>pframe_pipe_wq_v);
+ if (wait_event_interruptible(gtpframe_pipe_wq,
+ GTP_PIPE_PEEK >= 0)
+ == -ERESTARTSYS)
+ return -ERESTARTSYS;
+ }
+#undef GTP_PIPE_PEEK
+
+ /* Head. */
+ rbe = ring_buffer_consume(gtp_frame, cpu, &ts, NULL);
+ if (rbe == NULL)
+ goto recheck;
+ next = ring_buffer_event_data(rbe);
+ if (FID(next) != FID_HEAD)
+ goto recheck;
+ tmp16p = (int16_t *)gtp_realloc(grs, 2, 0);
+ if (!tmp16p)
+ return -ENOMEM;
+ *tmp16p = (int16_t)*(ULONGEST *)(next + FID_SIZE);
+ grs->size += 2;
+ /* This part is for the data_size. */
+ if (gtp_realloc(grs, 4, 0) == NULL)
+ return -ENOMEM;
+ grs->size += 4;
+ data_size = 0;
+
+ {
+ /* Handle $cpu_id and $clock. */
+ struct gtp_frame_var *vr;
+ char frame[GTP_FRAME_VAR_SIZE];
+
+
+ vr = (struct gtp_frame_var *) (frame + FID_SIZE);
+ vr->num = GTP_VAR_CLOCK_ID;
+ vr->val = ts;
+ if (gtp_frame2file_v(grs, &data_size, frame))
+ return -ENOMEM;
+ vr->num = GTP_VAR_CPU_ID;
+ vr->val = cpu;
+ if (gtp_frame2file_v(grs, &data_size, frame))
+ return -ENOMEM;
+ }
+
+ while (1) {
+#define GTP_PIPE_CONSUME (rbe = ring_buffer_consume(gtp_frame, cpu,
NULL, NULL))
+ GTP_PIPE_CONSUME;
+ if (rbe == NULL) {
+ if (!gtp_start)
+ break;
+
+ atomic_inc(>pframe_pipe_wq_v);
+ if (wait_event_interruptible(gtpframe_pipe_wq,
+ GTP_PIPE_CONSUME
+ != NULL)
+ == -ERESTARTSYS)
+ return -ERESTARTSYS;
+ continue;
+ }
+#undef GTP_PIPE_CONSUME
+ next = ring_buffer_event_data(rbe);
+
+ switch (FID(next)) {
+ case FID_REG:
+ if (gtp_frame2file_r(grs, &data_size, next))
+ return -ENOMEM;
+ break;
+
+ case FID_MEM:
+ if (gtp_frame2file_m(grs, &data_size, next))
+ return -ENOMEM;
+ break;
+
+ case FID_VAR:
+ if (gtp_frame2file_v(grs, &data_size, next))
+ return -ENOMEM;
+ break;
+
+ case FID_HEAD:
+ case FID_END:
+ goto while_out;
+ break;
+ }
+ }
+while_out:
+ /* Set the data_size. */
+ memcpy(grs->buf + grs->size - data_size - 4, &data_size, 4);
+
+ return 0;
+}
+
+static ssize_t
+gtpframe_pipe_read(struct file *file, char __user *buf, size_t size,
+ loff_t *ppos)
+{
+ ssize_t ret = -ENOMEM;
+ struct gtpframe_pipe_s *gps = file->private_data;
+ loff_t entry_offset;
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtpframe_pipe_read: size=%u *ppos=%lld\n",
+ size, *ppos);
+#endif
+
+ if (gps->grs->buf == NULL) {
+ gps->grs->real_size = 200;
+ gps->grs->buf = vmalloc(gps->grs->real_size);
+ if (!gps->grs->buf)
+ goto out;
+ gps->grs->size = 0;
+ } else if (*ppos < gps->begin
+ || *ppos >= (gps->begin + gps->grs->size)) {
+ gps->grs->size = 0;
+
+ if (gps->llseek_move) {
+ if (clear_user(buf, size)) {
+ ret = -EFAULT;
+ goto out;
+ }
+ gps->begin = 0;
+ gps->llseek_move = 0;
+ ret = size;
+ goto out;
+ }
+ }
+
+ if (gps->grs->size == 0) {
+ if (*ppos == 0) {
+
+ if (gtp_frame_file_header(gps->grs, 1))
+ goto out;
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtpframe_pipe_read: Get header.\n");
+#endif
+ } else {
+ ret = gtpframe_pipe_get_entry(gps->grs);
+ if (ret < 0)
+ goto out;
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtpframe_pipe_read: Get entry.\n");
+#endif
+ }
+ gps->begin = *ppos;
+ }
+
+#ifdef GTP_DEBUG
+ printk(GTP_DEBUG "gtpframe_pipe_read: gps->begin=%lld "
+ "gps->grs->size=%u\n",
+ gps->begin, gps->grs->size);
+#endif
+
+ entry_offset = *ppos - gps->begin;
+ ret = size;
+ if (entry_offset + size > gps->grs->size)
+ ret = gps->grs->size - entry_offset;
+ if (copy_to_user(buf, gps->grs->buf + entry_offset, ret)) {
+ ret = -EFAULT;
+ goto out;
+ }
+ *ppos += ret;
+
+out:
+ return ret;
+}
+
+static loff_t
+gtpframe_pipe_llseek(struct file *file, loff_t offset, int origin)
+{
+ struct gtpframe_pipe_s *gps = file->private_data;
+ loff_t ret = default_llseek(file, offset, origin);
+
+ if (ret < 0)
+ return ret;
+
+ if (ret >= gps->begin + gps->grs->size && gps->begin)
+ gps->llseek_move = 1;
+
+ return ret;
+}
+#endif
+
+static const struct file_operations gtp_operations = {
+ .owner = THIS_MODULE,
+ .open = gtp_open,
+ .release = gtp_release,
+ .unlocked_ioctl = gtp_ioctl,
+ .compat_ioctl = gtp_ioctl,
+ .read = gtp_read,
+ .write = gtp_write,
+ .poll = gtp_poll,
+};
+
+static const struct file_operations gtpframe_operations = {
+ .owner = THIS_MODULE,
+ .open = gtpframe_open,
+ .release = gtpframe_release,
+ .read = gtpframe_read,
+ .llseek = default_llseek,
+};
+
+#ifndef GTP_FRAME_SIMPLE
+static const struct file_operations gtpframe_pipe_operations = {
+ .owner = THIS_MODULE,
+ .open = gtpframe_pipe_open,
+ .release = gtpframe_pipe_release,
+ .read = gtpframe_pipe_read,
+ .llseek = gtpframe_pipe_llseek,
+};
+#endif
+
+struct dentry *gtp_dir;
+struct dentry *gtpframe_dir;
+#ifndef GTP_FRAME_SIMPLE
+struct dentry *gtpframe_pipe_dir;
+#endif
+
+static int __init gtp_init(void)
+{
+ int ret = -ENOMEM;
+
+ gtp_gtp_pid = -1;
+ gtp_gtp_pid_count = 0;
+ gtp_gtpframe_pid = -1;
+ gtp_gtpframe_pid_count = 0;
+#ifndef GTP_FRAME_SIMPLE
+ gtp_gtpframe_pipe_pid = -1;
+ gtp_gtpframe_pipe_pid_count = 0;
+#endif
+ gtp_list = NULL;
+ gtp_read_ack = 0;
+ gtp_rw_bufp = NULL;
+ gtp_rw_size = 0;
+ gtp_start = 0;
+ gtp_disconnected_tracing = 0;
+ gtp_circular = 0;
+ gtp_var_list = GTP_VAR_LIST_FIRST;
+ gtp_var_head = GTP_VAR_SPECIAL_MIN;
+ gtp_var_tail = GTP_VAR_SPECIAL_MAX;
+ gtp_var_array = NULL;
+ current_gtp_var = NULL;
+ gtp_frame = NULL;
+ gtp_frame_current_num = -1;
+ gtp_frame_current_tpe = 0;
+#ifdef GTP_FRAME_SIMPLE
+ gtp_frame_r_start = NULL;
+ gtp_frame_w_start = NULL;
+ gtp_frame_end = NULL;
+ gtp_frame_current = NULL;
+ gtp_frame_is_circular = 0;
+#else
+ {
+ int cpu;
+
+ for_each_online_cpu(cpu)
+ gtp_frame_iter[cpu] = NULL;
+ }
+ gtp_frame_current_cpu = 0;
+ atomic_set(>pframe_pipe_wq_v, 0);
+ gtpframe_pipe_is_running = 0;
+#endif
+ atomic_set(>p_frame_create, 0);
+ gtp_rw_count = 0;
+ gtp_frame_count = 0;
+ current_gtp = NULL;
+ current_gtp_action = NULL;
+ current_gtp_src = NULL;
+ gtpro_list = NULL;
+ gtp_frame_file = NULL;
+ gtp_frame_file_size = 0;
+ gtp_dir = NULL;
+ gtpframe_dir = NULL;
+#ifndef GTP_FRAME_SIMPLE
+ gtpframe_pipe_dir = NULL;
+#endif
+ {
+ int cpu;
+
+ gtp_cpu_number = 0;
+ for_each_online_cpu(cpu) {
+ if (cpu > gtp_cpu_number)
+ gtp_cpu_number = cpu;
+ }
+ gtp_cpu_number++;
+ }
+ gtp_start_last_errno = 0;
+ gtp_start_ignore_error = 0;
+
+ gtp_wq = create_singlethread_workqueue("gtpd");
+ if (gtp_wq == NULL)
+ goto out;
+
+ gtp_dir = debugfs_create_file("gtp", S_IFIFO | S_IRUSR | S_IWUSR, NULL,
+ NULL, >p_operations);
+ if (gtp_dir == NULL || gtp_dir == ERR_PTR(-ENODEV)) {
+ gtp_dir = NULL;
+ goto out;
+ }
+ gtpframe_dir = debugfs_create_file("gtpframe", S_IFIFO | S_IRUSR, NULL,
+ NULL, >pframe_operations);
+ if (gtpframe_dir == NULL || gtpframe_dir == ERR_PTR(-ENODEV)) {
+ gtpframe_dir = NULL;
+ goto out;
+ }
+#ifndef GTP_FRAME_SIMPLE
+ gtpframe_pipe_dir = debugfs_create_file("gtpframe_pipe",
+ S_IFIFO | S_IRUSR, NULL, NULL,
+ >pframe_pipe_operations);
+ if (gtpframe_pipe_dir == NULL
+ || gtpframe_pipe_dir == ERR_PTR(-ENODEV)) {
+ gtpframe_pipe_dir = NULL;
+ goto out;
+ }
+#endif
+
+ ret = 0;
+out:
+ if (ret < 0) {
+ if (gtp_wq)
+ destroy_workqueue(gtp_wq);
+
+ if (gtp_dir)
+ debugfs_remove_recursive(gtp_dir);
+ if (gtpframe_dir)
+ debugfs_remove_recursive(gtpframe_dir);
+#ifndef GTP_FRAME_SIMPLE
+ if (gtpframe_pipe_dir)
+ debugfs_remove_recursive(gtpframe_pipe_dir);
+#endif
+ }
+
+ return ret;
+}
+
+static void __exit gtp_exit(void)
+{
+ if (gtp_dir)
+ debugfs_remove_recursive(gtp_dir);
+ if (gtpframe_dir)
+ debugfs_remove_recursive(gtpframe_dir);
+#ifndef GTP_FRAME_SIMPLE
+ if (gtpframe_pipe_dir)
+ debugfs_remove_recursive(gtpframe_pipe_dir);
+#endif
+
+ gtp_gdbrsp_qtstop();
+ gtp_gdbrsp_qtinit();
+ if (gtp_frame) {
+#ifdef GTP_FRAME_SIMPLE
+ vfree(gtp_frame);
+#else
+ ring_buffer_free(gtp_frame);
+#endif
+ gtp_frame = NULL;
+ }
+
+ destroy_workqueue(gtp_wq);
+}
+
+module_init(gtp_init)
+module_exit(gtp_exit)
+
+MODULE_AUTHOR("Hui Zhu <teawater@...il.com>");
+MODULE_LICENSE("GPL");
--- /dev/null
+++ b/scripts/getgtprsp.pl
@@ -0,0 +1,102 @@
+#!/usr/bin/perl
+
+# This script to get the GDB tracepoint RSP package and save it
+# to ./gtpstart and ./gtpstop file.
+# GPL
+# Copyright(C) Hui Zhu (teawater@...il.com), 2010
+
+binmode STDIN, ":raw";
+$| = 1;
+
+$status = 0;
+$circular = 0;
+$var_count = 0;
+
+while (1) {
+ sysread STDIN, $c, 1 or next;
+ if ($c eq '') {
+ next;
+ } elsif ($c eq '+' || $c eq '-') {
+ $c = '';
+ }
+
+ sysread STDIN, $line, 1024 or next;
+ print '+';
+ $line = $c.$line;
+
+ open(LOG, ">>./log");
+ print LOG $line."\n";
+ close (LOG);
+
+ if ($status == 0) {
+ if ($line eq '$?#3f') {
+ print '$S05#b8';
+ } elsif ($line eq '$g#67') {
+ print '$00000000#80';
+ } elsif ($line =~ /^\$m/ || $line =~ /^\$p/) {
+ print '$00000000#80';
+ } elsif ($line eq '$qTStatus#49') {
+ print '$T0;tnotrun:0;tframes:0;tcreated:0;tsize:';
+ print '500000;tfree:500000;circular:0;disconn:0#d1';
+ } elsif ($line eq '$QTBuffer:circular:1#f9') {
+ print '$OK#9a';
+ $circular = 1;
+ } elsif ($line eq '$QTBuffer:circular:0#f8') {
+ print '$OK#9a';
+ $circular = 0;
+ } elsif ($line eq '$QTStop#4b') {
+ print '$OK#9a';
+ } elsif ($line =~ /^\$qSupported/) {
+ print '$ConditionalTracepoints+;TracepointSource+#1b';
+ } elsif ($line eq '$QTinit#59') {
+ $status = 1;
+ open(STARTFILE, ">./gtpstart");
+ print STARTFILE '$QTDisconnected:1#e3'."\n";
+ if ($circular) {
+ print STARTFILE '$QTBuffer:circular:1#f9'."\n";
+ } else {
+ print STARTFILE '$QTBuffer:circular:0#f8'."\n";
+ }
+ } elsif ($line eq '$qTfV#81') {
+ print '$8:0:1:64756d705f737461636b#f6';
+ } elsif ($line eq '$qTsV#8e') {
+ if ($var_count == 0) {
+ print '$7:0:1:7072696e746b5f666f726d6174#9b';
+ } elsif ($var_count == 1) {
+ print '$6:8:1:7072696e746b5f6c6576656c#3a';
+ } elsif ($var_count == 2) {
+ print '$5:0:1:7072696e746b5f746d70#28';
+ } elsif ($var_count == 3) {
+ print '$4:0:1:6370755f6964#f3';
+ } elsif ($var_count == 4) {
+ print '$3:0:1:636c6f636b#e1';
+ } elsif ($var_count == 5) {
+ print '$2:0:1:63757272656e745f7468726561';
+ print '645f696e666f#1f';
+ } elsif ($var_count == 6) {
+ print '$1:0:1:63757272656e745f7461736b#c7';
+ } else {
+ print '$l#6c';
+ }
+ $var_count++;
+ } else {
+ print '$#00';
+ }
+ }
+
+ if ($status == 1) {
+ print '$OK#9a';
+
+ print STARTFILE $line."\n";
+
+ if ($line eq '$QTStart#b3') {
+ $status = 0;
+
+ close(STARTFILE);
+
+ open(STOPFILE, ">./gtpstop");
+ print STOPFILE '$QTStop#4b'."\n";
+ close(STOPFILE);
+ }
+ }
+}
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists