[Bug breakpoints/10112] New: do not stop on function

[Bug breakpoints/10112] New: do not stop on function

[dushistov at mail dot ru]

here simple example:

#include <stdio.h>

void f(int gwk, double gsec)
{
        while (gsec < 0.0) {
                gwk--;
                gsec += 604800.0;
        }
        while (gsec >= 604800.0) {
                gwk++;
                gsec -= 604800.0;
        }

        printf("done\n");
}

int main()
{
        f(1322, 555198.9375);
        return 0;
}
-------------------------------------

gcc -ggdb -Wall test.c -o test_gdb && gdb test_gdb
GNU gdb 6.8
Copyright (C) 2008 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "i686-pc-linux-gnu"...
(gdb) br f
Breakpoint 1 at 0x80483f8: file test.c, line 6.
(gdb) r
Starting program: /var/tmp/test_gdb 
f was called

Program exited normally.
(gdb) 

As you can see function "f" was called, but gdb do not stop on it,

-- 
           Summary: do not stop on function
           Product: gdb
           Version: 6.8
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: breakpoints
        AssignedTo: unassigned at sourceware dot org
        ReportedBy: dushistov at mail dot ru
                CC: gdb-prs at sourceware dot org


http://sourceware.org/bugzilla/show_bug.cgi?id=10112

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug breakpoints/10112] do not stop on function

[pedro at codesourcery dot com]

------- Additional Comments From pedro at codesourcery dot com  2009-04-29 11:48 -------
Subject: Re:  New: do not stop on function

On Wednesday 29 April 2009 12:38:48, dushistov at mail dot ru wrote:
> here simple example:
> 
> #include <stdio.h>
> 
> void f(int gwk, double gsec)
> {
>         while (gsec < 0.0) {
>                 gwk--;
>                 gsec += 604800.0;
>         }
>         while (gsec >= 604800.0) {
>                 gwk++;
>                 gsec -= 604800.0;
>         }
> 
>         printf("done\n");
> }
> 
> int main()
> {
>         f(1322, 555198.9375);
>         return 0;
> }
> -------------------------------------
> 
> gcc -ggdb -Wall test.c -o test_gdb && gdb test_gdb

Most likely, the function wasn't called at all.  The
compiler probably inlined it in main.  You can confirm
that by looking at the disassembly of test_gdb.

Try recompiling, but this time, disabling optimizations:
 gcc -O0 -ggdb -Wall test.c -o test_gdb && gdb test_gdb
      ^^



-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=10112

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug breakpoints/10112] do not stop on function

[pedro at codesourcery dot com]

------- Additional Comments From pedro at codesourcery dot com  2009-04-29 11:49 -------
*** Bug 10113 has been marked as a duplicate of this bug. ***

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=10112

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug breakpoints/10112] do not stop on function

[dushistov at mail dot ru]

------- Additional Comments From dushistov at mail dot ru  2009-04-29 11:54 -------
(In reply to comment #1)
> > gcc -ggdb -Wall test.c -o test_gdb && gdb test_gdb
> 
> Most likely, the function wasn't called at all.  The
> compiler probably inlined it in main.  You can confirm
> that by looking at the disassembly of test_gdb.
> 
> Try recompiling, but this time, disabling optimizations:
>  gcc -O0 -ggdb -Wall test.c -o test_gdb && gdb test_gdb
>       ^^
> 

I don't think so, because of from documentation of gcc:
`-O0'
     Reduce compilation time and make debugging produce the expected
     results.  This is the default.

so it should be -O0 if no optimization parameters was used,
any way:

>gcc -O0 -ggdb -Wall test.c -o test_gdb && gdb test_gdb
GNU gdb 6.8
Copyright (C) 2008 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "i686-pc-linux-gnu"...
(gdb) br f
Breakpoint 1 at 0x80483f8: file test.c, line 6.
(gdb) r
Starting program: /var/tmp/test_gdb 
f was called

Program exited normally.
(gdb) 

the problem is place of breakpoint,
(gdb) info breakpoints
Num     Type           Disp Enb Address    What
1       breakpoint     keep y   0x080483f8 in f at test.c:6

(gdb) li f
1       #include <stdio.h>
2
3       void f(int gwk, double gsec)
4       {
5               while (gsec < 0.0) {
6                       gwk--;


so as you can see I put breakpoin of "f", but gdb put breakpoint
into while loop, and never happened, because of this example gsec >= 0



-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=10112

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug breakpoints/10112] do not stop on function

[pedro at codesourcery dot com]

------- Additional Comments From pedro at codesourcery dot com  2009-04-29 12:06 -------
Subject: Re:  do not stop on function

On Wednesday 29 April 2009 12:54:50, dushistov at mail dot ru wrote:

> (gdb) info breakpoints
> Num     Type           Disp Enb Address    What
> 1       breakpoint     keep y   0x080483f8 in f at test.c:6
> 
> (gdb) li f
> 1       #include <stdio.h>
> 2
> 3       void f(int gwk, double gsec)
> 4       {
> 5               while (gsec < 0.0) {
> 6                       gwk--;
> 
> 
> so as you can see I put breakpoin of "f", but gdb put breakpoint
> into while loop, and never happened, because of this example gsec >= 0

Ah, from your original post, I interpreted line 6 as
the `while' line...

> here simple example:
>                          <<< line 1
> #include <stdio.h>

Anyway, this is likely to be a debug info problem.  What's your gcc
version?  I just tried this with gcc version 4.2.4
(Ubuntu 4.2.4-1ubuntu3), and gdb 6.8, and it works correctly here:

(gdb) b f
Breakpoint 1 at 0x4004a8: file pr.c, line 5.
(gdb) r
Starting program: /home/pedro/test_gdb

Breakpoint 1, f (gwk=1322, gsec=555198.9375) at pr.c:5
5               while (gsec < 0.0) {
(gdb)


Also, please try a recent GDB snapshot.  I believe a couple
of related issues have been addressed since gdb 6.8 was out.



-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=10112

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug breakpoints/10112] do not stop on function

[dushistov at mail dot ru]

------- Additional Comments From dushistov at mail dot ru  2009-04-29 12:09 -------
(In reply to comment #4)
> Anyway, this is likely to be a debug info problem.  What's your gcc
> version?  

>gcc --version
gcc (Gentoo 4.3.2-r3 p1.6, pie-10.1.5) 4.3.2



> Also, please try a recent GDB snapshot.  I believe a couple
> of related issues have been addressed since gdb 6.8 was out.
> 
I will try. 



-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=10112

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug breakpoints/10112] do not stop on function

[dushistov at mail dot ru]

------- Additional Comments From dushistov at mail dot ru  2009-10-21 18:02 -------
I try recent version of gdb, bug still here:
gcc (Gentoo 4.3.2-r3 p1.6, pie-10.1.5) 4.3.2
GNU gdb (Gentoo 7.0 p1) 7.0

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=10112

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug breakpoints/10112] do not stop on function

[ppluzhnikov at google dot com]

------- Additional Comments From ppluzhnikov at google dot com  2009-10-21 19:27 -------
(In reply to comment #6)
> I try recent version of gdb, bug still here:
> gcc (Gentoo 4.3.2-r3 p1.6, pie-10.1.5) 4.3.2
> GNU gdb (Gentoo 7.0 p1) 7.0

Please verify that your .gdbinit is not interfering in unexpected way, by
running 'gdb -nx ./test_gdb'

Assuming that's not it, please supply output from:

 (gdb) info break
 (gdb) disassemble/m f

and

 readelf -w test_gdb    # probably best supplied as an attachment


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=10112

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug breakpoints/10112] do not stop on function

[dushistov at mail dot ru]

------- Additional Comments From dushistov at mail dot ru  2009-10-22 17:13 -------
Current directory is /home/evgeniy/projects/examples/gdb_not_stop_10112/

warning: Can not parse XML syscalls information; XML support was disabled at compile time.
GNU gdb (Gentoo 7.0 p1) 7.0
Copyright (C) 2009 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "i686-pc-linux-gnu".
For bug reporting instructions, please see:
<http://bugs.gentoo.org/>...
Reading symbols from /home/evgeniy/projects/examples/gdb_not_stop_10112/test...done.
(gdb) br f
Breakpoint 1 at 0x80483f8: file test.c, line 6.
(gdb) info break
Num     Type           Disp Enb Address    What
1       breakpoint     keep y   0x080483f8 in f at test.c:6
(gdb) disassemble/m f
Dump of assembler code for function f:
4	{
0x080483e4 <f+0>:	push   %ebp
0x080483e5 <f+1>:	mov    %esp,%ebp
0x080483e7 <f+3>:	sub    $0x18,%esp
0x080483ea <f+6>:	mov    0xc(%ebp),%eax
0x080483ed <f+9>:	mov    %eax,-0x8(%ebp)
0x080483f0 <f+12>:	mov    0x10(%ebp),%eax
0x080483f3 <f+15>:	mov    %eax,-0x4(%ebp)
0x080483f6 <f+18>:	jmp    0x804840a <f+38>

5	        while (gsec < 0.0) {
0x0804840a <f+38>:	fldl   -0x8(%ebp)
0x0804840d <f+41>:	fldz   
0x0804840f <f+43>:	fucomip %st(1),%st
0x08048411 <f+45>:	fstp   %st(0)
0x08048413 <f+47>:	ja     0x80483f8 <f+20>
0x08048415 <f+49>:	jmp    0x8048429 <f+69>

6	                gwk--;
0x080483f8 <f+20>:	subl   $0x1,0x8(%ebp)

7	                gsec += 604800.0;
0x080483fc <f+24>:	fldl   -0x8(%ebp)
0x080483ff <f+27>:	fldl   0x8048548
0x08048405 <f+33>:	faddp  %st,%st(1)
0x08048407 <f+35>:	fstpl  -0x8(%ebp)

8	        }
9	        while (gsec >= 604800.0) {
0x08048429 <f+69>:	fldl   -0x8(%ebp)
0x0804842c <f+72>:	fldl   0x8048548
0x08048432 <f+78>:	fxch   %st(1)
0x08048434 <f+80>:	fucomip %st(1),%st
0x08048436 <f+82>:	fstp   %st(0)
0x08048438 <f+84>:	jae    0x8048417 <f+51>

10	                gwk++;
0x08048417 <f+51>:	addl   $0x1,0x8(%ebp)

11	                gsec -= 604800.0;
0x0804841b <f+55>:	fldl   -0x8(%ebp)
0x0804841e <f+58>:	fldl   0x8048548
0x08048424 <f+64>:	fsubrp %st,%st(1)
0x08048426 <f+66>:	fstpl  -0x8(%ebp)

12	        }
13	
14	        printf("done\n");
0x0804843a <f+86>:	movl   $0x8048540,(%esp)
0x08048441 <f+93>:	call   0x80482fc <puts@plt>

15	}
0x08048446 <f+98>:	leave  
0x08048447 <f+99>:	ret    

End of assembler dump.
(gdb) 

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=10112

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug breakpoints/10112] do not stop on function

[dushistov at mail dot ru]

------- Additional Comments From dushistov at mail dot ru  2009-10-22 17:15 -------
Created an attachment (id=4309)
 --> (http://sourceware.org/bugzilla/attachment.cgi?id=4309&action=view)
readelf -w output


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=10112

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug breakpoints/10112] do not stop on function

[ppluzhnikov at google dot com]

------- Additional Comments From ppluzhnikov at google dot com  2009-10-22 17:47 -------
This sounds very familiar:

  Extended opcode 2: set Address to 0x80483e4
  Special opcode 8: advance Address by 0 to 0x80483e4 and Line by 3 to 4
  Advance PC by constant 17 to 0x80483f5
GDB sets breakpoint here ==>
  Special opcode 49: advance Address by 3 to 0x80483f8 and Line by 2 to 6
  Special opcode 62: advance Address by 4 to 0x80483fc and Line by 1 to 7
but should have set it here instead ==>
  Special opcode 199: advance Address by 14 to 0x804840a and Line by -2 to 5

I am about 95% certain this was fixed on the GCC side.
FWIW, in my (patched) GCC-4.3.1, the line table looks like this:

  Extended opcode 2: set Address to 0x8048260
  Special opcode 8: advance Address by 0 to 0x8048260 and Line by 3 to 4
  Advance PC by constant 17 to 0x8048271
This is the entry which causes GDB to work correctly ==>
  Special opcode 20: advance Address by 1 to 0x8048272 and Line by 1 to 5
  Special opcode 34: advance Address by 2 to 0x8048274 and Line by 1 to 6
  Special opcode 62: advance Address by 4 to 0x8048278 and Line by 1 to 7
  Special opcode 199: advance Address by 14 to 0x8048286 and Line by -2 to 5

And the corresponding disassembly:

(gdb) disas/m f
Dump of assembler code for function f:
4       {
   0x08048260 <+0>:     push   %ebp
   0x08048261 <+1>:     mov    %esp,%ebp
   0x08048263 <+3>:     sub    $0x18,%esp
   0x08048266 <+6>:     mov    0xc(%ebp),%eax
   0x08048269 <+9>:     mov    %eax,-0x8(%ebp)
   0x0804826c <+12>:    mov    0x10(%ebp),%eax
   0x0804826f <+15>:    mov    %eax,-0x4(%ebp)

5               while (gsec < 0.0) {
   0x08048272 <+18>:    jmp    0x8048286 <f+38>
   0x08048286 <+38>:    fldl   -0x8(%ebp)
   0x08048289 <+41>:    fldz   
   0x0804828b <+43>:    fucomip %st(1),%st
   0x0804828d <+45>:    fstp   %st(0)
   0x0804828f <+47>:    ja     0x8048274 <f+20>
   0x08048291 <+49>:    jmp    0x80482a5 <f+69>

6                       gwk--;
   0x08048274 <+20>:    subl   $0x1,0x8(%ebp)

7                       gsec += 604800.0;
   0x08048278 <+24>:    fldl   -0x8(%ebp)
   0x0804827b <+27>:    fldl   0x80483c8
   0x08048281 <+33>:    faddp  %st,%st(1)
   0x08048283 <+35>:    fstpl  -0x8(%ebp)

8               }


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=10112

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug breakpoints/10112] do not stop on function

[pedro at codesourcery dot com]

------- Additional Comments From pedro at codesourcery dot com  2010-08-14 02:49 -------
Is there anything to do on this PR then?


-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |WAITING


http://sourceware.org/bugzilla/show_bug.cgi?id=10112

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug breakpoints/10112] do not stop on function

[dushistov at mail dot ru]

http://sourceware.org/bugzilla/show_bug.cgi?id=10112

Evgeniy Dushistov <dushistov at mail dot ru> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |RESOLVED
         Resolution|                            |FIXED

--- Comment #12 from Evgeniy Dushistov <dushistov at mail dot ru> 2011-05-23 13:12:07 UTC ---
>Is there anything to do on this PR then?

I test with gcc 4.4.5 and gdb 7.2.
And cann't reproduce this bug, so fixed.

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