[Bug gdb/31896] New: -exec-next skips statements

[Bug gdb/31896] New: -exec-next skips statements

[dmitry.neverov at jetbrains dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31896

            Bug ID: 31896
           Summary: -exec-next skips statements
           Product: gdb
           Version: 14.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: gdb
          Assignee: unassigned at sourceware dot org
          Reporter: dmitry.neverov at jetbrains dot com
  Target Milestone: ---

I'm debugging the following code, the at line1 (v == 0):

  if (v) { // line1
    // branch1
  } else if (predicate()) { // line2
    // branch2
  }
  else 
  {
    // branch3, line3
  }

When I issue the `next` command, execution stops at line3 instead of
line2. If I change the code and add a newline before the 'else if',
`next` stops at line2 as expected:

  if (condition1) { // line1
    // branch1
  }
  else if (predicate()) { // line2
    // branch2
  }
  else 
  {
    // branch3, line3
  }

It looks like it happens because the last jump instruction in first
branch has the line2.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug gdb/31896] -exec-next skips statements

[tromey at sourceware dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=31896

Tom Tromey <tromey at sourceware dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |tromey at sourceware dot org

--- Comment #1 from Tom Tromey <tromey at sourceware dot org> ---
Normally this kind of thing turns out to be caused
by the compiler.  You can try examining the line
table to see.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug gdb/31896] -exec-next skips statements

[dmitry.neverov at jetbrains dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=31896

--- Comment #2 from Dmitry Neverov <dmitry.neverov at jetbrains dot com> ---
Proper steps to reproduce on linux/aarch64 with main.cpp:

#include <iostream>

bool predicate(int v) {
    return v == 7;
}

void f(int v) {
    if (v) { // break here
        std::cout << "branch1\n";
    } else if(predicate(v)) {
        std::cout << "branch2\n";
    }
    else {
        std::cout << "branch3\n";
    }
}

int main() {
    f(0);
    return 0;
}


$ clang++-15 -g -O0 main.cpp

$ gdb a.out

(gdb) b main.cpp:8
Breakpoint 1 at 0xa10: file main.cpp, line 8.
(gdb) run
Starting program: a.out
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/aarch64-linux-gnu/libthread_db.so.1".
Breakpoint 1, f (v=0) at main.cpp:8
8           if (v) {
(gdb) next
14              std::cout << "branch3\n";
(gdb) quit

I can reach the predicate() call if I use nexti (gdb stops at
0x0000aaaaaaaa0a38).

(gdb) info line *0x0000aaaaaaaa0a38
Line 10 of "main.cpp" starts at address 0xaaaaaaaa0a38 <_Z1fi+68> and ends at
0xaaaaaaaa0a3c <_Z1fi+72>.

Line 10 starts earlier than this address and this confuses gdb:

(gdb) info line 10
Line 10 of "main.cpp" starts at address 0xaaaaaaaa0a34 <_Z1fi+64> and ends at
0xaaaaaaaa0a38 <_Z1fi+68>.

Is it considered a compiler bug?

-- 
You are receiving this mail because:
You are on the CC list for the bug.