[Bug tdep/15420] New: Thread support broken on 64-bit Solaris/x86

[Bug tdep/15420] New: Thread support broken on 64-bit Solaris/x86

[ro at TechFak dot Uni-Bielefeld.DE]

http://sourceware.org/bugzilla/show_bug.cgi?id=15420

             Bug #: 15420
           Summary: Thread support broken on 64-bit Solaris/x86
           Product: gdb
           Version: HEAD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: tdep
        AssignedTo: unassigned@sourceware.org
        ReportedBy: ro@TechFak.Uni-Bielefeld.DE
                CC: brobecker@adacore.com
    Classification: Unclassified
              Host: amd64-pc-solaris2.1[01]
            Target: amd64-pc-solaris2.1[01]
             Build: amd64-pc-solaris2.1[01]


This report is about gdb 7.6, but bugzilla doesn't yet provide that version.

When I start gdb, configured on Solaris 10/x86 with

configure CFLAGS='-g -O2 -m64' --prefix=/vol/gcc --disable-nls --without-python

I get this warning:

[GDB will not be able to debug user-mode threads: ld.so.1: gdb-7.6: fatal:
relocation error: file /lib/64/libthread_db.so.1: symbol ps_lgetxregsize:
referenced symbol not found]

This is a regression from gdb 7.5 where this worked and the function (required
by libthread_db.so.1) was provided by the gdb binary.  This is obviously due to
this change:

2012-09-27  Joel Brobecker  <brobecker@adacore.com>

        * sol-thread.c (ps_lgetxregsize, ps_lgetxregs, ps_lsetxregs):
        Move these functions to sparc-sol-thread.c.
        * sparc-sol-thread.c: New file.
        * configure.ac: Add sparc-sol-thread.o to CONFIG_OBS and
        sparc-sol-thread.c to CONFIG_SRCS for sparc-solaris native
        configurations.
        * configure: Regenerate.

The functions are incorrectly taken to be sparc-specific.  If I add
sparc-sol-thread.c to the Solaris/x86 build by editing configure, gdb starts
again as expected.

  Rainer

-- 
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.

[Bug tdep/15420] Thread support broken on 64-bit Solaris/x86

[brobecker at gnat dot com]

http://sourceware.org/bugzilla/show_bug.cgi?id=15420

Joel Brobecker <brobecker at gnat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
                 CC|                            |brobecker at gnat dot com
         AssignedTo|unassigned at sourceware    |brobecke at sourceware dot
                   |dot org                     |org

--- Comment #1 from Joel Brobecker <brobecker at gnat dot com> 2013-05-01 06:43:33 UTC ---
Sorry for the breakage. Will fix.

-- 
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.

[Bug tdep/15420] Thread support broken on 64-bit Solaris/x86

[brobecker at gnat dot com]

http://sourceware.org/bugzilla/show_bug.cgi?id=15420

--- Comment #2 from Joel Brobecker <brobecker at gnat dot com> 2013-05-06 11:35:14 UTC ---
Rainer,

The Solaris 10 man page is absolutely explicit about these routines:

  SPARC Only
     ps_lgetxregsize(),ps_lgetxregs(),   andps_lsetxregs()    are
     SPARC-specific. They do not need to be defined by a control-
     ling process on  non-SPARC  architecture.  ps_lgetxregsize()
     returns  in *xregsize the size of the architecture-dependent
     extra state registers. ps_lgetxregs() gets the  extra  state
     registers, and ps_lsetxregs() sets them.

This may be obsolete info, but I am testing on ia32-solaris, and I am not
seeing the problem either. I wonder if the 'x' might mean 'cross' meaning that
the routines are only used when in 64bit mode...

I will send a patch...

-- 
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.

[Bug tdep/15420] Thread support broken on 64-bit Solaris/x86

[ro at CeBiTec dot Uni-Bielefeld.DE]

http://sourceware.org/bugzilla/show_bug.cgi?id=15420

--- Comment #3 from Rainer Orth <ro at CeBiTec dot Uni-Bielefeld.DE> 2013-05-06 11:48:08 UTC ---
> --- Comment #2 from Joel Brobecker <brobecker at gnat dot com> 2013-05-06 11:35:14 UTC ---
> Rainer,
>
> The Solaris 10 man page is absolutely explicit about these routines:
>
>   SPARC Only
>      ps_lgetxregsize(),ps_lgetxregs(),   andps_lsetxregs()    are
>      SPARC-specific. They do not need to be defined by a control-
>      ling process on  non-SPARC  architecture.  ps_lgetxregsize()
>      returns  in *xregsize the size of the architecture-dependent
>      extra state registers. ps_lgetxregs() gets the  extra  state
>      registers, and ps_lsetxregs() sets them.

I'm seeing this only in the Solaris 9 man page.  On Solaris 10 (Update
11, man page dated 13 Sep 2010), I have

     ps_lgetxregsize(), ps_lgetxregs(),  and  ps_lsetxregs()  are
     system dependent. ps_lgetxregsize() returns in *xregsize the
     size of the architecture-dependent  extra  state  registers.
     ps_lgetxregs()   gets   the   extra   state  registers,  and
     ps_lsetxregs() sets them. On systems  that  do  not  support
     extra   state   registers,   these   functions  will  return
     PS_NOXREGS.

> This may be obsolete info, but I am testing on ia32-solaris, and I am not
> seeing the problem either. I wonder if the 'x' might mean 'cross' meaning that
> the routines are only used when in 64bit mode...

I suppose the x means extra, as in proc(4):

  xregs
     Extra state registers.  The  extra  state  register  set  is
     architecture  dependent;  this  file  is empty if the system
     does not support extra state registers. If the file is  non-
     empty,  it  contains  an architecture dependent structure of
     type prxregset_t, defined in <procfs.h>, with the values  of
     the  lwp's extra state registers. If the lwp is not stopped,
     all register values are undefined. See also the PCSXREG con-
     trol operation, below.

> I will send a patch...

Fine, thanks.

    Rainer

-- 
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.

[Bug tdep/15420] Thread support broken on 64-bit Solaris/x86

[brobecker at gnat dot com]

http://sourceware.org/bugzilla/show_bug.cgi?id=15420

--- Comment #4 from Joel Brobecker <brobecker at gnat dot com> 2013-05-06 12:12:03 UTC ---
Patch sent.

For the record, the system I copied the man page from is running x86-solaris
2.10.

-- 
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.

[Bug tdep/15420] Thread support broken on 64-bit Solaris/x86

[cvs-commit at gcc dot gnu.org]

http://sourceware.org/bugzilla/show_bug.cgi?id=15420

--- Comment #5 from cvs-commit at gcc dot gnu.org <cvs-commit at gcc dot gnu.org> 2013-05-10 12:10:22 UTC ---
CVSROOT:    /cvs/src
Module name:    src
Changes by:    brobecke@sourceware.org    2013-05-10 12:10:20

Modified files:
    gdb            : ChangeLog configure configure.ac sol-thread.c 
Removed files:
    gdb            : sparc-sol-thread.c 

Log message:
    move sparc-sol-thread.c back into sol-thread.c.

    The routines in sparc-sol-thread used to be SPARC-specific (and
    documented as such in the ptrace man page), and therefore hosting them
    in a sparc-specific file made sense.  However, newer versions of
    Solaris now use those callbacks (Solaris 10 Update 10, apparently),
    and thus the note about these callbacks being specific to SPARC
    was removed.

    So this patch deletes sparc-sol-thread.c and moves the code back
    inside sol-thread.c.

    gdb/ChangeLog:

    PR tdep/15420:
    * sol-thread.c (ps_lgetxregsize, ps_lgetxregs, ps_lsetxregs):
    New functions, directly copied from sparc-sol-thread.c.
    * sparc-sol-thread.c: Delete.
    * configure.ac: Remove code handling sparc-solaris-thread.c.
    * configure: Regenerate.

Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gdb/ChangeLog.diff?cvsroot=src&r1=1.15540&r2=1.15541
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gdb/configure.diff?cvsroot=src&r1=1.389&r2=1.390
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gdb/configure.ac.diff?cvsroot=src&r1=1.201&r2=1.202
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gdb/sol-thread.c.diff?cvsroot=src&r1=1.100&r2=1.101
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gdb/sparc-sol-thread.c.diff?cvsroot=src&r1=1.2&r2=NONE

-- 
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.

[Bug tdep/15420] Thread support broken on 64-bit Solaris/x86

[cvs-commit at gcc dot gnu.org]

http://sourceware.org/bugzilla/show_bug.cgi?id=15420

--- Comment #6 from cvs-commit at gcc dot gnu.org <cvs-commit at gcc dot gnu.org> 2013-05-10 12:30:07 UTC ---
CVSROOT:    /cvs/src
Module name:    src
Branch:     gdb_7_6-branch
Changes by:    brobecke@sourceware.org    2013-05-10 12:30:05

Modified files:
    gdb            : ChangeLog configure configure.ac sol-thread.c 
Removed files:
    gdb            : sparc-sol-thread.c 

Log message:
    move sparc-sol-thread.c back into sol-thread.c.

    The routines in sparc-sol-thread used to be SPARC-specific (and
    documented as such in the ptrace man page), and therefore hosting them
    in a sparc-specific file made sense.  However, newer versions of
    Solaris now use those callbacks (Solaris 10 Update 10, apparently),
    and thus the note about these callbacks being specific to SPARC
    was removed.

    So this patch deletes sparc-sol-thread.c and moves the code back
    inside sol-thread.c.

    gdb/ChangeLog:

    PR tdep/15420:
    * sol-thread.c (ps_lgetxregsize, ps_lgetxregs, ps_lsetxregs):
    New functions, directly copied from sparc-sol-thread.c.
    * sparc-sol-thread.c: Delete.
    * configure.ac: Remove code handling sparc-solaris-thread.c.
    * configure: Regenerate.

Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gdb/ChangeLog.diff?cvsroot=src&only_with_tag=gdb_7_6-branch&r1=1.15260.2.51&r2=1.15260.2.52
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gdb/configure.diff?cvsroot=src&only_with_tag=gdb_7_6-branch&r1=1.385.2.1&r2=1.385.2.2
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gdb/configure.ac.diff?cvsroot=src&only_with_tag=gdb_7_6-branch&r1=1.197.2.1&r2=1.197.2.2
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gdb/sol-thread.c.diff?cvsroot=src&only_with_tag=gdb_7_6-branch&r1=1.94&r2=1.94.2.1
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gdb/sparc-sol-thread.c.diff?cvsroot=src&only_with_tag=gdb_7_6-branch&r1=1.2&r2=NONE

-- 
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.

[Bug tdep/15420] Thread support broken on 64-bit Solaris/x86

[brobecker at gnat dot com]

http://sourceware.org/bugzilla/show_bug.cgi?id=15420

Joel Brobecker <brobecker at gnat dot com> changed:

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

--- Comment #7 from Joel Brobecker <brobecker at gnat dot com> 2013-05-10 12:37:02 UTC ---
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.

[Bug tdep/15420] Thread support broken on 64-bit Solaris/x86

[mehmetgelisin at aol dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=15420

Mehmet gelisin <mehmetgelisin at aol dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mehmetgelisin at aol dot com

--- Comment #8 from Mehmet gelisin <mehmetgelisin at aol dot com> ---
have started experiencing a similar crash when we switched to g++ 7.1 with
c++17 semantics.  When gdb tries to print a "static constexpr", it triggers a
"error reading variable: Missing ELF symbol" exception and then crashes at the
same place as the original reporter. This seems to happen with every release at
least from gdb 7.11. I also built the git master and could reproduce the
problem as well.

It's not entirely clear that this is the same exact bug but the crash backtrace
is completely similar

I have a simple reproducer: https://komiya-dental.com/

$ cat main.cpp 
struct A {
    static constexpr const char *a = "a";
    void foo() { }
};

int main(void) {
    A a;
    a.foo();
}

Compile with g++ 7.1 with C++17 semantics: g++-7 -std=c++17 -g -o main main.cpp

Now run it with gdb:

$ gdb main
GNU gdb (Debian 7.11.1-2~bpo8+1) 7.11.1
(...)
(gdb) b A::foo
Breakpoint 1 at 0x40055e: file main.cpp, line 6.
(gdb) r http://www.iu-bloomington.com/
Starting program: main 

Breakpoint 1, A::foo (this=0x7fffffffe09f) at main.cpp:6
6           void foo() { }
(gdb) print *this
Segmentation fault (core dumped) https://www.webb-dev.co.uk/ 

If I patch the source with the workaround mentioned by Thilo (i.e set the ptr
to NULL in the catch block and call cp_print_static_field() only if v is not
NULL), it does not crash:
(gdb) print *this
$1 = {static a = <error reading variable: Missing ELF symbol "A::a".>}
 https://waytowhatsnext.com/  
If I compile the program above with the default c++ semantics(c++ 14), I get 
(gdb) print *this
$1 = {static a = <optimized out>}

With the current git master (8.0.50.20170801-git), this is the backtrace I get
for the gdb crash: http://www.acpirateradio.co.uk/ 

Program terminated with signal SIGSEGV, Segmentation fault.
#0  value_entirely_covered_by_range_vector (value=0x0, ranges=0x90) at
value.c:401
401       if (value->lazy)
(gdb) bt
#0  value_entirely_covered_by_range_vector (value=0x0, ranges=0x90) at
value.c:401
#1  0x0000000000563282 in cp_print_static_field (options=<optimized out>,
http://www.logoarts.co.uk/ recurse=<optimized out>, stream=<optimized out>,
val=<optimized out>, type=<optimized out>) at cp-valprint.c:640
#2  cp_print_value_fields (type=0x0, real_type=0x90, real_type@entry=0xf1cd00,
offset=offset@entry=0, address=0, address@entry=140737488347295,
http://www.slipstone.co.uk/  stream=0xf194c0, recurse=16998736,
recurse@entry=0, val=0xf8fc00, options=0x7fffffffda20, dont_print_vb=0x0,
dont_print_statmem=0)
    at cp-valprint.c:335 http://embermanchester.uk/ 
#3  0x00000000005639ca in cp_print_value_fields_rtti (type=<optimized out>,
type@entry=0xf1cd00, valaddr=0xf6a330 "", offset=offset@entry=0,
http://connstr.net/ 
 address=140737488347295, stream=0xf194c0, recurse=0, val=0xf8fc00,
options=0x7fffffffda20, dont_print_vb=0x0, dont_print_statmem=0)
    at cp-valprint.c:456 http://joerg.li/
#4  0x000000000054bf52 in c_val_print_struct (type=0xf1cd00, valaddr=<optimized
out>, embedded_offset=0, address=<optimized out>, stream=<optimized out>,
recurse=<optimized out>, original_value=0xf8fc00, options=0x7fffffffda20) at
c-valprint.c:412 http://www.jopspeech.com/
#5  0x000000000054c46f in c_val_print (type=0xf1cd00, embedded_offset=0,
address=0, stream=0xf194c0, recurse=0, original_value=0xf8fc00,
options=0x7fffffffda20) at c-valprint.c:530 http://www.wearelondonmade.com/
#6  0x0000000000699270 in val_print (type=0x0, type@entry=0xf1cd00,
embedded_offset=0, address=15846656, address@entry=140737488347295,
http://www.compilatori.com/ stream=stream@entry=0xf194c0, recurse=32767,
recurse@entry=0, val=val@entry=0xf8fc00, options=0x7fffffffdae0,
language=0x843b80  http://www-look-4.com/ <cplus_language_defn>)

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

[Bug tdep/15420] Thread support broken on 64-bit Solaris/x86

[mervegunesli at aol dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=15420

Merve Gunesli <mervegunesli at aol dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mervegunesli at aol dot com

--- Comment #9 from Merve Gunesli <mervegunesli at aol dot com> ---
http://www.acpirateradio.co.uk/category/travel/
http://www.slipstone.co.uk/category/health/
http://embermanchester.uk/tech/nvidia-and-samsung/
http://joerg.li/category/travel/
http://www.jopspeech.com/computers/latest-car-deals/
http://www.compilatori.com/category/computers/

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

[Bug tdep/15420] Thread support broken on 64-bit Solaris/x86

[mark at klomp dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=15420

Mark Wielaard <mark at klomp dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mark at klomp dot org

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