[Bug testsuite/30704] New: [gdb/testsuite, i686] FAIL: gdb.base/catch-syscall.exp: multiple targets: i386 vs i386:x86-64: insert catch syscall on syscall 1 -- write on i386:x86-64

[Bug testsuite/30704] New: [gdb/testsuite, i686] FAIL: gdb.base/catch-syscall.exp: multiple targets: i386 vs i386:x86-64: insert catch syscall on syscall 1 -- write on i386:x86-64

[vries at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=30704

            Bug ID: 30704
           Summary: [gdb/testsuite, i686] FAIL:
                    gdb.base/catch-syscall.exp: multiple targets: i386 vs
                    i386:x86-64: insert catch syscall on syscall 1 --
                    write on i386:x86-64
           Product: gdb
           Version: HEAD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: testsuite
          Assignee: unassigned at sourceware dot org
          Reporter: vries at gcc dot gnu.org
  Target Milestone: ---

I build gdb on native i686-pc-linux-gnu with --enable-targets=all, and ran the
testsuite, and ran into:
...
FAIL: gdb.base/catch-syscall.exp: multiple targets: i386 vs i386:x86-64: insert
catch syscall on syscall 1 -- write on i386:x86-64
FAIL: gdb.base/catch-syscall.exp: multiple targets: sparc vs sparc:v9: insert
catch syscall on syscall 108 -- setresuid on sparc:v9
...

In more detail:
...
(gdb) PASS: gdb.base/catch-syscall.exp: multiple targets: i386 vs i386:x86-64:
set architecture i386:x86-64
catch syscall 1^M
The feature 'catch syscall' is not supported on this architecture yet.^M
(gdb) FAIL: gdb.base/catch-syscall.exp: multiple targets: i386 vs i386:x86-64:
insert catch syscall on syscall 1 -- write on i386:x86-64
...

I tracked this down to amd64-linux-tdep.c not being compiled, so
amd64_linux_init_abi is not called, which would call
amd64_linux_init_abi_common, which does:
...
  set_gdbarch_get_syscall_number (gdbarch,
                                  amd64_linux_get_syscall_number);
...

I looked in Makefile.in, and found amd64-linux-tdep.o listed in
ALL_64_TARGET_OBS:
...
# All target-dependent objects files that require 64-bit CORE_ADDR              
# (used with --enable-targets=all --enable-64-bit-bfd).                         
ALL_64_TARGET_OBS = \
...
so presumably I can make the test-case pass by building with
--enable-64-bit-bfd.

I'm a bit confused that it's valid to set the arch to i386:x86-64 without
--enable-64-bit-bfd, but perhaps that's ok.

So I guess we need to update the test-case.

Note that we do have:
...
$ gdb -q -batch -ex "set arch i386:x86-64"
warning: A handler for the OS ABI "GNU/Linux" is not built into this
configuration
of GDB.  Attempting to continue with the default i386:x86-64 settings.
...
The target architecture is set to "i386:x86-64".

So perhaps we can require the set arch command to issue no warnings.

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

[Bug testsuite/30704] [gdb/testsuite, i686] FAIL: gdb.base/catch-syscall.exp: multiple targets: i386 vs i386:x86-64: insert catch syscall on syscall 1 -- write on i386:x86-64

[vries at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=30704

--- Comment #1 from Tom de Vries <vries at gcc dot gnu.org> ---
Interestingly, "set arch aarch64" fails:
...
(gdb) set architecture aarch64^M
Undefined item: "aarch64".^M
(gdb) UNSUPPORTED: gdb.base/catch-syscall.exp: multiple targets: aarch64 vs
arm: set architecture aarch64
set architecture arm^M
The target architecture is set to "arm".^M
(gdb) 
...

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

[Bug testsuite/30704] [gdb/testsuite, i686] FAIL: gdb.base/catch-syscall.exp: multiple targets: i386 vs i386:x86-64: insert catch syscall on syscall 1 -- write on i386:x86-64

[vries at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=30704

--- Comment #2 from Tom de Vries <vries at gcc dot gnu.org> ---
Created attachment 15025
  --> https://sourceware.org/bugzilla/attachment.cgi?id=15025&action=edit
Tentative patch

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

[Bug testsuite/30704] [gdb/testsuite, i686] FAIL: gdb.base/catch-syscall.exp: multiple targets: i386 vs i386:x86-64: insert catch syscall on syscall 1 -- write on i386:x86-64

[tromey at sourceware dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=30704

Tom Tromey <tromey at sourceware dot org> changed:

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

--- Comment #3 from Tom Tromey <tromey at sourceware dot org> ---
(In reply to Tom de Vries from comment #0)

> I'm a bit confused that it's valid to set the arch to i386:x86-64 without
> --enable-64-bit-bfd, but perhaps that's ok.

I find this surprising as well.

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

[Bug testsuite/30704] [gdb/testsuite, i686] FAIL: gdb.base/catch-syscall.exp: multiple targets: i386 vs i386:x86-64: insert catch syscall on syscall 1 -- write on i386:x86-64

[vries at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=30704

--- Comment #4 from Tom de Vries <vries at gcc dot gnu.org> ---
(In reply to Tom de Vries from comment #0)
> so presumably I can make the test-case pass by building with
> --enable-64-bit-bfd.

Confirmed, the test-case passes fully after a rebuild with --enable-64-bit-bfd.

Now also this works:
...
(gdb) set architecture aarch64^M
The target architecture is set to "aarch64".^M
(gdb)
...

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