Re: [PATCH] Add NetBSD/aarch64 gdbserver support

[Luis Machado <luis.machado@linaro.org>]

Hi,

On 10/5/20 9:01 PM, Kamil Rytarowski wrote:
> The support is on par with NetBSD/amd64, thus GPR works,
> single step and software breakpoint are operational, and the
> SVR4 r_debug integration is functional.
> 
> gdbserver/ChangeLog:
> 
>          * netbsd-aarch64-low.cc: Add.
>          * Makefile.in (SFILES): Register "netbsd-aarch64-low.c".
>          * configure.srv: Add aarch64*-*-netbsd*.
> ---
>   gdbserver/ChangeLog             |   6 ++
>   gdbserver/Makefile.in           |   1 +
>   gdbserver/configure.srv         |   6 ++
>   gdbserver/netbsd-aarch64-low.cc | 113 ++++++++++++++++++++++++++++++++
>   4 files changed, 126 insertions(+)
>   create mode 100644 gdbserver/netbsd-aarch64-low.cc
> 
> diff --git a/gdbserver/ChangeLog b/gdbserver/ChangeLog
> index d667a203e67..5fd71b6ee71 100644
> --- a/gdbserver/ChangeLog
> +++ b/gdbserver/ChangeLog
> @@ -1,3 +1,9 @@
> +2020-10-01  Kamil Rytarowski  <n54@gmx.com>
> +
> +	* netbsd-aarch64-low.cc: Add.
> +	* Makefile.in (SFILES): Register "netbsd-aarch64-low.c".
> +	* configure.srv: Add aarch64*-*-netbsd*.
> +
>   2020-10-01  Kamil Rytarowski  <n54@gmx.com>
> 
>   	* netbsd-i386-low.cc: Add.
> diff --git a/gdbserver/Makefile.in b/gdbserver/Makefile.in
> index 8b80acebdf1..442a34b381f 100644
> --- a/gdbserver/Makefile.in
> +++ b/gdbserver/Makefile.in
> @@ -193,6 +193,7 @@ SFILES = \
>   	$(srcdir)/linux-x86-low.cc \
>   	$(srcdir)/linux-xtensa-low.cc \
>   	$(srcdir)/mem-break.cc \
> +	$(srcdir)/netbsd-aarch64-low.cc \
>   	$(srcdir)/netbsd-amd64-low.cc \
>   	$(srcdir)/netbsd-i386-low.cc \
>   	$(srcdir)/netbsd-low.cc \
> diff --git a/gdbserver/configure.srv b/gdbserver/configure.srv
> index c4fe4ca81ee..cd25b1574e0 100644
> --- a/gdbserver/configure.srv
> +++ b/gdbserver/configure.srv
> @@ -61,6 +61,12 @@ case "${gdbserver_host}" in
>   			ipa_obj="${ipa_obj} linux-aarch64-tdesc-ipa.o"
>   			ipa_obj="${ipa_obj} arch/aarch64-ipa.o"
>   			;;
> +  aarch64*-*-netbsd*)	srv_regobj=""
> +			srv_tgtobj="netbsd-low.o netbsd-aarch64-low.o fork-child.o"
> +			srv_tgtobj="${srv_tgtobj} nat/fork-inferior.o"
> +			srv_tgtobj="${srv_tgtobj} nat/netbsd-nat.o"
> +			srv_tgtobj="${srv_tgtobj} arch/aarch64-insn.o arch/aarch64.o"
> +			;;
>     arm*-*-linux*)	srv_tgtobj="$srv_linux_obj linux-arm-low.o"
>   			srv_tgtobj="$srv_tgtobj linux-arm-tdesc.o"
>   			srv_tgtobj="$srv_tgtobj linux-aarch32-low.o"
> diff --git a/gdbserver/netbsd-aarch64-low.cc b/gdbserver/netbsd-aarch64-low.cc
> new file mode 100644
> index 00000000000..d0e485649d9
> --- /dev/null
> +++ b/gdbserver/netbsd-aarch64-low.cc
> @@ -0,0 +1,113 @@
> +/* Copyright (C) 2020 Free Software Foundation, Inc.
> +
> +   This file is part of GDB.
> +
> +   This program is free software; you can redistribute it and/or modify
> +   it under the terms of the GNU General Public License as published by
> +   the Free Software Foundation; either version 3 of the License, or
> +   (at your option) any later version.
> +
> +   This program is distributed in the hope that it will be useful,
> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> +   GNU General Public License for more details.
> +
> +   You should have received a copy of the GNU General Public License
> +   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
> +
> +#include <sys/types.h>
> +#include <sys/ptrace.h>
> +#include <limits.h>
> +
> +#include "server.h"
> +#include "netbsd-low.h"
> +#include "arch/aarch64.h"
> +#include "arch/aarch64-insn.h"
> +#include "tdesc.h"
> +
> +/* The fill_function for the general-purpose register set.  */
> +
> +static void
> +netbsd_aarch64_fill_gregset (struct regcache *regcache, char *buf)
> +{
> +  struct reg *r = (struct reg *) buf;
> +
> +#define netbsd_aarch64_collect_gp(regnum, fld) do {		\
> +    collect_register (regcache, regnum, &r->fld);		\
> +  } while (0)

Do we really need the above instead of using the for loop below?

> +
> +  for (size_t i = 0; i < 31; i++)
> +    netbsd_aarch64_collect_gp (AARCH64_X0_REGNUM + i, r_reg[i]);

A matter of taste, but we could define a reasonably-named constant for 
"31" instead of hardcoding it here.

> +
> +  netbsd_aarch64_collect_gp (AARCH64_SP_REGNUM, r_sp);
> +  netbsd_aarch64_collect_gp (AARCH64_PC_REGNUM, r_pc);
> +}
> +
> +/* The store_function for the general-purpose register set.  */
> +
> +static void
> +netbsd_aarch64_store_gregset (struct regcache *regcache, const char *buf)
> +{
> +  struct reg *r = (struct reg *) buf;
> +
> +#define netbsd_aarch64_supply_gp(regnum, fld) do {		\
> +    supply_register (regcache, regnum, &r->fld);		\
> +  } while(0)

Same as above. Do we really need this construct?

> +
> +  for (size_t i = 0; i < 31; i++)
> +    netbsd_aarch64_supply_gp (AARCH64_X0_REGNUM + i, r_reg[i]);

Same suggestion to replace the hardcoded constant with a define.

> +
> +  netbsd_aarch64_supply_gp (AARCH64_SP_REGNUM, r_sp);
> +  netbsd_aarch64_supply_gp (AARCH64_PC_REGNUM, r_pc);
> +}
> +
> +/* Description of all the aarch64-netbsd register sets.  */
> +
> +static const struct netbsd_regset_info netbsd_target_regsets[] =
> +{
> +  /* General Purpose Registers.  */
> +  {PT_GETREGS, PT_SETREGS, sizeof (struct reg),
> +  netbsd_aarch64_fill_gregset, netbsd_aarch64_store_gregset},
> +  /* End of list marker.  */
> +  {0, 0, -1, NULL, NULL }
> +};
> +
> +/* NetBSD target op definitions for the aarch64 architecture.  */
> +
> +class netbsd_aarch64_target : public netbsd_process_target
> +{
> +protected:
> +  const netbsd_regset_info *get_regs_info () override;
> +
> +  void low_arch_setup () override;
> +};
> +
> +/* Return the information to access registers.  */
> +
> +const netbsd_regset_info *
> +netbsd_aarch64_target::get_regs_info ()
> +{
> +  return netbsd_target_regsets;
> +}
> +
> +/* Architecture-specific setup for the current process.  */
> +
> +void
> +netbsd_aarch64_target::low_arch_setup ()
> +{
> +  target_desc *tdesc
> +    = aarch64_create_target_description (0, false);

Did you choose to go with the standard core registers with CPSR and no 
TPIDR?

Otherwise this looks good and I don't have any more comments.