[Bug c++/29243] New: infinite loop in generic_value_print

[Bug c++/29243] New: infinite loop in generic_value_print

[roland at gnu dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=29243

            Bug ID: 29243
           Summary: infinite loop in generic_value_print
           Product: gdb
           Version: HEAD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: c++
          Assignee: unassigned at sourceware dot org
          Reporter: roland at gnu dot org
  Target Milestone: ---

Created attachment 14142
  --> https://sourceware.org/bugzilla/attachment.cgi?id=14142&action=edit
x86_64-linux-gnu target assembly source

The source to reproduce this is quite small. However, GCC 12 does not emit
DWARF that causes the issue in GDB.
Current ToT Clang does.  I've attached the .s file produced by Clang; turning
that into an object that you can load into gdb should reproduce the same
problem.

```
// clang++ -std=c++17 -O2 -g3 -gdwarf-5 -shared -o foo.so foo.cc                
// gdb foo.so                                                                   
// (gdb) i func                                                                 

struct S { bool x, y; };                                                        

namespace {                                                                     

template <auto M>                                                               
constexpr bool kFoo = true;                                                     

} //                                                                            

bool t() { return kFoo<&S::x>; }                                                

bool f() { return kFoo<&S::y>; }   
```

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

[Bug c++/29243] infinite loop in generic_value_print

[roland at gnu dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=29243

Roland McGrath <roland at gnu dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at sourceware dot org   |tromey at sourceware dot org
             Status|NEW                         |ASSIGNED

--- Comment #1 from Roland McGrath <roland at gnu dot org> ---
Commit 6b5a7bc768570edc15532759969cd9bff3f02c28 seems suspect here.
Tom, can you take a look?

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

[Bug c++/29243] infinite loop in generic_value_print

[tromey at sourceware dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=29243

--- Comment #2 from Tom Tromey <tromey at sourceware dot org> ---
(In reply to Roland McGrath from comment #0)
> Current ToT Clang does.  I've attached the .s file produced by Clang;
> turning that into an object that you can load into gdb should reproduce the
> same problem.

gas doesn't like this .s file.
Can you just attach a .o.

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

[Bug c++/29243] infinite loop in generic_value_print

[roland at gnu dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=29243

--- Comment #3 from Roland McGrath <roland at gnu dot org> ---
Created attachment 14143
  --> https://sourceware.org/bugzilla/attachment.cgi?id=14143&action=edit
gas-compatible assembly source

This is a version of the same .s file that gas accepts.

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

[Bug c++/29243] infinite loop in generic_value_print

[roland at gnu dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=29243

--- Comment #4 from Roland McGrath <roland at gnu dot org> ---
Created attachment 14144
  --> https://sourceware.org/bugzilla/attachment.cgi?id=14144&action=edit
.o file

Here is the .o file from that .s file.

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

[Bug c++/29243] infinite loop in generic_value_print

[tromey at sourceware dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=29243

Tom Tromey <tromey at sourceware dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |13.1

--- Comment #5 from Tom Tromey <tromey at sourceware dot org> ---
https://sourceware.org/pipermail/gdb-patches/2022-October/192804.html

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

[Bug c++/29243] infinite loop in generic_value_print

[tromey at sourceware dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=29243

Tom Tromey <tromey at sourceware dot org> changed:

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

--- Comment #6 from Tom Tromey <tromey at sourceware dot org> ---
Fixed.

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

[Bug c++/29243] infinite loop in generic_value_print

[cvs-commit at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=29243

--- Comment #7 from cvs-commit at gcc dot gnu.org <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Tom Tromey <tromey@sourceware.org>:

https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;h=86430497337968e6163aef370c6312e7b5ed6504

commit 86430497337968e6163aef370c6312e7b5ed6504
Author: Tom Tromey <tromey@adacore.com>
Date:   Tue Oct 18 11:32:52 2022 -0600

    Fix regression in pointer-to-member printing

    PR c++/29243 points out that "info func" on a certain C++ executable
    will cause an infinite loop in gdb.

    I tracked this down to a bug introduced by commit 6b5a7bc76 ("Handle
    member pointers directly in generic_value_print").  Before this
    commit, the C++ code to print a member pointer would wind up calling
    value_print_scalar_formatted; but afterward it simply calls
    generic_value_print and gets into a loop.

    This patch restores the previous behavior and adds a regression test.

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