[Bug testsuite/4794] backtrace of yyy_func3 and print_stack of yyy_func3 fails on ppc64.

[Bug testsuite/4794] backtrace of yyy_func3 and print_stack of yyy_func3 fails on ppc64.

[mjw at redhat dot com]

http://sourceware.org/bugzilla/show_bug.cgi?id=4794

Mark Wielaard <mjw at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
                 CC|                            |mjw at redhat dot com
         Resolution|                            |FIXED

--- Comment #5 from Mark Wielaard <mjw at redhat dot com> 2011-11-29 09:59:30 UTC ---
Both user and kernel backtraces should now work with systemtap git.

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

[Bug testsuite/4794] backtrace of yyy_func3 and print_stack of yyy_func3 fails on ppc64.

[srinivasa at in dot ibm dot com]

------- Additional Comments From srinivasa at in dot ibm dot com  2008-01-17 12:40 -------
I have investigated further on this. 
  Stack generating function __stp_stack_print() looks similar dump_stack() and 
I found that, this problem is present when we call dump_stack() from kprobe 
handler.

Hence this problem could be due to, 
1) we might have stored the ip somewhere else in the stack and reading the 
different value. ( Hence regs->nip at the first frame shows the correct value 
and value read from the stack shows inavlid value)

2) We might have overwritten the ip value in that stack.

dump_stack() from kprobe handler.
=============================================
 LR = .yyy_func2+0x10/0x28 [systemtap_test_module2]
[c0000000355dbab0] [0000000000000080] 0x80 (unreliable)
[c0000000355dbb30] [d0000000006e006c] .yyy_func1+0x10/0x28 
[systemtap_test_module2]
[c0000000355dbbb0] [d000000000790164] .stm_write_cmd+0xbc/0x440 
[systemtap_test_module1]
[c0000000355dbc40] [c000000000141cbc] .proc_reg_write+0xc0/0x100
[c0000000355dbcf0] [c0000000000f4dc0] .vfs_write+0xf8/0x1c4
[c0000000355dbd90] [c0000000000f57c0] .sys_write+0x4c/0x8c
[c0000000355dbe30] [c000000000008734] syscall_exit+0x0/0x40
========================================

stack generated by systemtap.
==========================================
 0xd0000000006e000c : .yyy_func3+0x0/0x28 
[systemtap_test_module2]0xc0000000355dbab0
[0xc0000000355db838] [0x00000000000004d2] 0x00000000000004d2  (unreliable)
[0xc0000000355dbab0] [0x0000000000000080] 0x0000000000000080
[0xc0000000355dbb30] [0xd0000000006e006c] 0xd0000000006e006c : .yyy_func1+0x10/
0x28 [systemtap_test_module2]
[0xc0000000355dbbb0] [0xd000000000790164] 0xd000000000790164
 : .stm_write_cmd+0xbc/0x0 [systemtap_test_module1]
[0xc0000000355dbc40] [0xc000000000141cbc] 0xc000000000141cbc
 : .proc_reg_write+0xc0/0x100 []
[0xc0000000355dbcf0] [0xc0000000000f4dc0] 0xc0000000000f4dc0 : .vfs_write+0xf8/
0x1c4 []
[0xc0000000355dbd90] [0xc0000000000f57c0] 0xc0000000000f57c0 : .sys_write+0x4c/
0x8c []
[0xc0000000355dbe30] [0xc000000000008734] 0xc000000000008734 : syscall_exit+0x0/
0x40 []
--- Exception: c00 at 0x000000000ff1a758
===============================

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4794

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug testsuite/4794] backtrace of yyy_func3 and print_stack of yyy_func3 fails on ppc64.

[srinivasa at in dot ibm dot com]

------- Additional Comments From srinivasa at in dot ibm dot com  2008-01-16 14:25 -------
Jim
 Iam sorry, I mean 2 frame information in stack is not "missing" , it is just 
invalid.

Thanks
 Srinivasa DS

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4794

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug testsuite/4794] backtrace of yyy_func3 and print_stack of yyy_func3 fails on ppc64.

[srinivasa at in dot ibm dot com]

------- Additional Comments From srinivasa at in dot ibm dot com  2008-01-16 14:17 -------
Jim
Since we are missing 2 frame information in backtrace output of ppc64(as shown 
below),I looked in to backtrace code(runtime/stack-ppc64.c) for ppc64 and got 
some doubts.
could you please clarify.
==============
backtrace from module("systemtap_test_module2").function("yyy_func3@/home/
systemtap/tmp/stap_testing_200801141002/src/testsuite/systemtap.cont:^M
 0xd000000000bf000c : .yyy_func3+0x0/0x28 [systemtap_test_module2]^M
[0xc00000005d8938e8] [0x00000000000004d2] 0x00000000000004d2 (unreliable)^M
[0xc00000005d893b60] [0xc00000005d893c60] 0xc00000005d893c60^M
[0xc00000005d893be0] [0xd000000000bf006c] 0xd000000000bf006c : .yyy_func1+0x10/
0x28 [systemtap_test_module2]^M
[0xc00000005d893c60] [0xd000000000a2014c] 0xd000000000a2014c
 : .stm_write_cmd+0xa4/0x0 [systemtap_test_module1]^M
[0xc00000005d893cf0] [0xc0000000000ed31c] 0xc0000000000ed31c : .vfs_write+0x118/
0x200 []^M
[0xc00000005d893d90] [0xc0000000000eda8c] 0xc0000000000eda8c : .sys_write+0x4c/
0x8c []^M
[0xc00000005d893e30] [0xc0000000000086a4] 0xc0000000000086a4 : syscall_exit+0x0/
0x40 []^M
==============================
 

1) while unwinding the stack, we are not starting from REGS_SP but one frame 
below from that, 
 In above case REGS_SP = 0xc00000005d893b60 but we unwind from 
0xc00000005d8938e8.

2) Even I tried unwinding from REGS_SP, but to my surprise, I found that IP 
value from the stack( ip = _sp[2]) is not same as REGS_IP and hence we are not 
seeing symbol for first 2 frames.

 


-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ananth at in dot ibm dot
                   |                            |com, jkenisto at us dot ibm
                   |                            |dot com


http://sourceware.org/bugzilla/show_bug.cgi?id=4794

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug testsuite/4794] backtrace of yyy_func3 and print_stack of yyy_func3 fails on ppc64.

[srinivasa at in dot ibm dot com]

------- Additional Comments From srinivasa at in dot ibm dot com  2007-09-17 09:38 -------
backtrace and print_stack of yyy_func4, backtrace and print_stack of
time_profile fails with this week snapshot(systemtap-20070915.tar.bz2)



=======================================================================
Running
/home/systemtap/tmp/stap_testing_200709170524/src/testsuite/systemtap.context/context.exp
...
Systemtap probe: begin^M
PASS: backtrace of begin probe
backtrace from
module("systemtap_test_module2").function("yyy_func3@/home/systemtap/tmp/stap_testing_20070917052
4/src/testsuite/systemtap.cont:^M
 0xd00000000076000c : .yyy_func3+0x0/0x28 [systemtap_test_module2]^M
[0xc00000000f52b838] [0x00000000000004d2] 0x00000000000004d2 (unreliable)^M
[0xc00000000f52bab0] [0x0000000000000020] 0x0000000000000020^M
[0xc00000000f52bb30] [0xd00000000076006c] 0xd00000000076006c :
.yyy_func1+0x10/0x28 [systemtap_test_module2]^M
[0xc00000000f52bbb0] [0xd000000000780164] 0xd000000000780164 :
.stm_write_cmd+0xbc/0x0 [systemtap_test_module1]^
M
[0xc00000000f52bc40] [0xc0000000001422d0] 0xc0000000001422d0 :
.proc_reg_write+0xc0/0x100 []^M
[0xc00000000f52bcf0] [0xc0000000000eac40] 0xc0000000000eac40 :
.vfs_write+0x120/0x1f0 []^M
[0xc00000000f52bd90] [0xc0000000000eb670] 0xc0000000000eb670 :
.sys_write+0x4c/0x8c []^M
[0xc00000000f52be30] [0xc000000000008534] 0xc000000000008534 :
syscall_exit+0x0/0x40 []^M
--- Exception: c00 at 0x000000000ff1a758^M
    LR =0x000000000feb63ec^M
--------^M
the call stack is 0xd00000000076000c 0x00000000000004d2 0x0000000000000020
0xd00000000076006c 0xd000000000780164
 0xc0000000001422d0 0xc0000000000^M
--------^M
 0xd00000000076000c : .yyy_func3+0x0/0x28 [systemtap_test_module2]^M
 0x00000000000004d2^M
 0x0000000000000020^M
 0xd00000000076006c : .yyy_func1+0x10/0x28 [systemtap_test_module2]^M
 0xd000000000780164 : .stm_write_cmd+0xbc/0x0 [systemtap_test_module1]^M
 0xc0000000001422d0 : .proc_reg_write+0xc0/0x100 []^M
 0x00000c0000000000^M
--------^M
backtrace from
module("systemtap_test_module2").function("yyy_func4@/home/systemtap/tmp/stap_testing_200709170524/src/testsuite/systemtap.cont:^M
Returning from: 0xd000000000760000 : .yyy_func4+0x0/0xc [systemtap_test_module2]^M
Returning to  : 0xd00000000076001c : .yyy_func3+0x10/0x28 [systemtap_test_module2]^M
[0xc00000000f52b7b8] [0x00000000000004d3] 0x00000000000004d3 (unreliable)^M
[0xc00000000f52ba30] [0xc00000000f52bb10] 0xc00000000f52bb10^M
[0xc00000000f52bab0] [0xd000000000760044] 0xd000000000760044 :
.yyy_func2+0x10/0x28 [systemtap_test_module2]^M
[0xc00000000f52bb30] [0xd00000000076006c] 0xd00000000076006c :
.yyy_func1+0x10/0x28 [systemtap_test_module2]^M
[0xc00000000f52bbb0] [0xd000000000780164] 0xd000000000780164 :
.stm_write_cmd+0xbc/0x0 [systemtap_test_module1]^M
[0xc00000000f52bc40] [0xc0000000001422d0] 0xc0000000001422d0 :
.proc_reg_write+0xc0/0x100 []^M
[0xc00000000f52bcf0] [0xc0000000000eac40] 0xc0000000000eac40 :
.vfs_write+0x120/0x1f0 []^M
[0xc00000000f52bd90] [0xc0000000000eb670] 0xc0000000000eb670 :
.sys_write+0x4c/0x8c []^M
[0xc00000000f52be30] [0xc000000000008534] 0xc000000000008534 :
syscall_exit+0x0/0x40 []^M
--- Exception: c00 at 0x000000000ff1a758^M
    LR =0x000000000feb63ec^M
--------^M
the return stack is 0xc00000000002ad10 0x00000000000004d3 0xc00000000f52bb10
0xd000000000760044 0xd00000000076006c 0xd000000000780164 0xc0000000001^M
--------^M
 0xc00000000002ad10 : kretprobe_trampoline+0x0/0x8 []^M
 0x00000000000004d3^M
 0xc00000000f52bb10^M
 0xd000000000760044 : .yyy_func2+0x10/0x28 [systemtap_test_module2]^M
 0xd00000000076006c : .yyy_func1+0x10/0x28 [systemtap_test_module2]^M
 0xd000000000780164 : .stm_write_cmd+0xbc/0x0 [systemtap_test_module1]^M
 0x00000c0000000001^M
--------^M
backtrace from timer.profile:^M
 0xc00000000000b28c : .local_irq_restore+0x3c/0x40 []^M
[0xc000000000673b68] [0x0000000000000001] 0x0000000000000001 (unreliable)^M
[0xc000000000673de0] [0xc000000000011904] 0xc000000000011904 :
.cpu_idle+0x108/0x1e0 []^M
[0xc000000000673e60] [0xc0000000003828c4] 0xc0000000003828c4 :
.rest_init+0x7c/0x94 []^M
[0xc000000000673ee0] [0xc0000000004f09d4] 0xc0000000004f09d4^M
[0xc000000000673f90] [0xc000000000382844] 0xc000000000382844 :
.start_here_common+0x54/0x58 []^M
--------^M
the profile stack is 0xc00000000000b28c 0x0000000000000001 0xc000000000011904
0xc0000000003828c4 0xc0000000004f09d4 0xc000000000382844^M
--------^M
 0xc00000000000b28c : .local_irq_restore+0x3c/0x40 []^M
 0x0000000000000001^M
 0xc000000000011904 : .cpu_idle+0x108/0x1e0 []^M
 0xc0000000003828c4 : .rest_init+0x7c/0x94 []^M
 0xc0000000004f09d4^M
 0xc000000000382844 : .start_here_common+0x54/0x58 []^M
--------^M
FAIL: backtrace of yyy_func3 (1)
FAIL: print_stack of yyy_func3 (1)
FAIL: backtrace of yyy_func4.return (1)
FAIL: print_stack of yyy_func4.return (1)
FAIL: backtrace of timer.profile (1)
FAIL: print_stack of timer.profile (1)
READY^M
yyy_int -1 200 300^M
yyy_int returns 499^M
PASS: integer function arguments
yyy_uint 4294967295 200 300^M
yyy_uint returns 499^M
PASS: unsigned function arguments
yyy_long -1 200 300^M
yyy_long returns 499^M
PASS: long function arguments
yyy_int64 -1 200 300^M
yyy_int64 returns 499^M
PASS: int64 function arguments
yyy_char a b c^M
yyy_char returns Q^M
PASS: char function arguments
yyy_str Hello-System-Tap^M
yyy_str returns XYZZY^M
PASS: string function arguments
READY^M
execname: echo^M
pexecname: expect^M
pid: 20647^M
ppid: 11635^M
tid: 20647^M
uid: 0^M
euid: 0^M
gid: 0^M
egid: 0^M
PASS: execname
PASS: pexecname
PASS: pid
PASS: ppid
PASS: tid
PASS: uid
PASS: euid
PASS: gid
PASS: egid
testcase
/home/systemtap/tmp/stap_testing_200709170524/src/testsuite/systemtap.context/context.exp
completed in 57 seconds
===============================================

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4794

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.