[PATCH] Create avr-linux-tdep.c (add reintroduce gdbarch_gdb_signal_{to,from}_target for that target)

[PATCH] Create avr-linux-tdep.c (add reintroduce gdbarch_gdb_signal_{to,from}_target for that target)

[Sergio Durigan Junior]

Hi,

Given the discussion on:

<http://sourceware.org/ml/gdb-patches/2013-08/msg00290.html>
<http://sourceware.org/ml/gdb-patches/2013-08/msg00296.html>

Here is the patch to add avr-linux-tdep.c.  This patch also reimplements
gdbarch_gdb_signal_{to,from}_target for AVR targets running on Linux.

Joern, could you please give it a try (I don't have access to AVR
targets here), and tell me what you think?

Thanks,

-- 
Sergio

2013-08-12  Sergio Durigan Junior  <sergiodj@redhat.com>

	* Makefile.in (ALL_TARGET_OBS): Add avr-linux-tdep.o.
	(ALLDEPFILES): Add avr-linux-tdep.c.
	* avr-linux-tdep.c: New file.
	* configure.tgt: Add match for avr-*-*linux*.

diff --git a/gdb/Makefile.in b/gdb/Makefile.in
index 9171940..0be4198 100644
--- a/gdb/Makefile.in
+++ b/gdb/Makefile.in
@@ -548,7 +548,7 @@ ALL_TARGET_OBS = \
 	armbsd-tdep.o arm-linux-tdep.o arm-symbian-tdep.o \
 	armnbsd-tdep.o armobsd-tdep.o \
 	arm-tdep.o arm-wince-tdep.o \
-	avr-tdep.o \
+	avr-linux-tdep.o avr-tdep.o \
 	bfin-linux-tdep.o bfin-tdep.o \
 	cris-tdep.o \
 	dicos-tdep.o \
@@ -1473,7 +1473,7 @@ ALLDEPFILES = \
 	amd64-sol2-tdep.c \
 	arm-linux-nat.c arm-linux-tdep.c arm-symbian-tdep.c arm-tdep.c \
 	armnbsd-nat.c armbsd-tdep.c armnbsd-tdep.c armobsd-tdep.c \
-	avr-tdep.c \
+	avr-linux-tdep.c avr-tdep.c \
 	bfin-linux-tdep.c bfin-tdep.c \
 	bsd-uthread.c bsd-kvm.c \
 	core-regset.c \
diff --git a/gdb/avr-linux-tdep.c b/gdb/avr-linux-tdep.c
new file mode 100644
index 0000000..e45ddb7
--- /dev/null
+++ b/gdb/avr-linux-tdep.c
@@ -0,0 +1,115 @@
+/* Target-dependent code for Atmel AVR, for GDB.
+
+   Copyright (C) 2013 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 "defs.h"
+#include "gdbarch.h"
+#include "osabi.h"
+#include "linux-tdep.h"
+
+/* This enum represents the signals' numbers on the AVR
+   architecture.  It just contains the signal definitions which are
+   different from the generic implementation.
+
+   It is derived from the file <arch/avr32/include/uapi/asm/signal.h>,
+   from the Linux kernel tree.  */
+
+enum
+  {
+    AVR_LINUX_SIGRTMIN = 32,
+    AVR_LINUX_SIGRTMAX = 63,
+  };
+
+/* Implementation of `gdbarch_gdb_signal_from_target', as defined in
+   gdbarch.h.  */
+
+static enum gdb_signal
+avr_linux_gdb_signal_from_target (struct gdbarch *gdbarch, int signal)
+{
+  if (signal >= AVR_LINUX_SIGRTMIN && signal <= AVR_LINUX_SIGRTMAX)
+    {
+      int offset = signal - AVR_LINUX_SIGRTMIN;
+
+      if (offset == 0)
+	return GDB_SIGNAL_REALTIME_32;
+      else
+	return (enum gdb_signal) (offset - 1
+				  + (int) GDB_SIGNAL_REALTIME_33);
+    }
+  else if (signal > AVR_LINUX_SIGRTMAX)
+    return GDB_SIGNAL_UNKNOWN;
+
+  return linux_gdb_signal_from_target (gdbarch, signal);
+}
+
+/* Implementation of `gdbarch_gdb_signal_to_target', as defined in
+   gdbarch.h.  */
+
+static int
+avr_linux_gdb_signal_to_target (struct gdbarch *gdbarch,
+				enum gdb_signal signal)
+{
+  switch (signal)
+    {
+    /* GDB_SIGNAL_REALTIME_32 is not continuous in <gdb/signals.def>,
+       therefore we have to handle it here.  */
+    case GDB_SIGNAL_REALTIME_32:
+      return AVR_LINUX_SIGRTMIN;
+
+    /* GDB_SIGNAL_REALTIME_64 is not valid on AVR.  */
+    case GDB_SIGNAL_REALTIME_64:
+      return -1;
+    }
+
+  /* GDB_SIGNAL_REALTIME_33 to _63 are continuous.
+     AVR does not have _64.  */
+  if (signal >= GDB_SIGNAL_REALTIME_33
+      && signal <= GDB_SIGNAL_REALTIME_63)
+    {
+      int offset = signal - GDB_SIGNAL_REALTIME_33;
+
+      return AVR_LINUX_SIGRTMIN + 1 + offset;
+    }
+
+  return linux_gdb_signal_to_target (gdbarch, signal);
+}
+
+\f
+
+/* Initialize ABI for AVR.  */
+
+static void
+avr_linux_init_abi (struct gdbarch_info info, struct gdbarch *gdbarch)
+{
+  linux_init_abi (info, gdbarch);
+
+  set_gdbarch_gdb_signal_from_target (gdbarch,
+				      avr_linux_gdb_signal_from_target);
+  set_gdbarch_gdb_signal_to_target (gdbarch,
+				    avr_linux_gdb_signal_to_target);
+}
+
+/* Silence -Wmissing-prototypes.  */
+extern initialize_file_ftype _initialize_avr_linux_tdep;
+
+void
+_initialize_avr_linux_tdep (void)
+{
+  gdbarch_register_osabi (bfd_arch_avr, 0, GDB_OSABI_LINUX,
+                          avr_linux_init_abi);
+}
diff --git a/gdb/configure.tgt b/gdb/configure.tgt
index 653ba2b..500c596 100644
--- a/gdb/configure.tgt
+++ b/gdb/configure.tgt
@@ -111,6 +111,12 @@ arm*-*-*)
 	gdb_sim=../sim/arm/libsim.a
 	;;
 
+avr-*-*linux*)
+	# Target: AVR Linux
+	gdb_target_obs="avr-tdep.o avr-linux-tdep.o linux-tdep.o"
+	gdb_sim=../sim/avr/libsim.a
+	;;
+
 avr-*-*)
 	# Target: AVR
 	gdb_target_obs="avr-tdep.o"

Re: [PATCH] Create avr-linux-tdep.c (add reintroduce gdbarch_gdb_signal_{to,from}_target for that target)

[Joern Rennecke]

Quoting Sergio Durigan Junior <sergiodj@redhat.com>:


> +avr-*-*linux*)
> +	# Target: AVR Linux
> +	gdb_target_obs="avr-tdep.o avr-linux-tdep.o linux-tdep.o"
> +	gdb_sim=../sim/avr/libsim.a
> +	;;
> +
>  avr-*-*)
>  	# Target: AVR
>  	gdb_target_obs="avr-tdep.o"
>

That is creating a new target.  So are users using a linux host supposed
to configure for this new target?  But why does it look like it's the
target that is running linux - there is no such thing for avr.
Shouldn't we rather autoconf the host signals?

Re: [PATCH] Create avr-linux-tdep.c (add reintroduce gdbarch_gdb_signal_{to,from}_target for that target)

[Sergio Durigan Junior]

On Monday, August 12 2013, Joern Rennecke wrote:

> Quoting Sergio Durigan Junior <sergiodj@redhat.com>:
>
>
>> +avr-*-*linux*)
>> +	# Target: AVR Linux
>> +	gdb_target_obs="avr-tdep.o avr-linux-tdep.o linux-tdep.o"
>> +	gdb_sim=../sim/avr/libsim.a
>> +	;;
>> +
>>  avr-*-*)
>>  	# Target: AVR
>>  	gdb_target_obs="avr-tdep.o"
>>
>
> That is creating a new target.

Yes.

> So are users using a linux host supposed
> to configure for this new target?  But why does it look like it's the
> target that is running linux - there is no such thing for avr.
> Shouldn't we rather autoconf the host signals?

These are not host signals, they are the target's.

Maybe I am confusing things here.  I saw
<arch/avr32/include/uapi/asm/signal.h> and I assumed one could run Linux
on AVR, but on a further investigation I saw that it only applies to
AVR32, right?  I'm definitely not an expert here, sorry about it.

But well, if that is the case, then I guess we can either (a) treat
AVR32 separately and make an avr32-linux-tdep.c for it, or (b) drop this
patch entirely.

WDYT?  Am I missing something here?

-- 
Sergio

Re: [PATCH] Create avr-linux-tdep.c (add reintroduce gdbarch_gdb_signal_{to,from}_target for that target)

[Pedro Alves]

On 08/12/2013 10:48 PM, Sergio Durigan Junior wrote:
> On Monday, August 12 2013, Joern Rennecke wrote:
> 
>> Quoting Sergio Durigan Junior <sergiodj@redhat.com>:
>>
>>
>>> +avr-*-*linux*)
>>> +	# Target: AVR Linux
>>> +	gdb_target_obs="avr-tdep.o avr-linux-tdep.o linux-tdep.o"
>>> +	gdb_sim=../sim/avr/libsim.a
>>> +	;;
>>> +
>>>  avr-*-*)
>>>  	# Target: AVR
>>>  	gdb_target_obs="avr-tdep.o"
>>>
>>
>> That is creating a new target.
> 
> Yes.
> 
>> So are users using a linux host supposed
>> to configure for this new target?  But why does it look like it's the
>> target that is running linux - there is no such thing for avr.
>> Shouldn't we rather autoconf the host signals?
> 
> These are not host signals, they are the target's.
> 
> Maybe I am confusing things here.  I saw
> <arch/avr32/include/uapi/asm/signal.h> and I assumed one could run Linux
> on AVR, but on a further investigation I saw that it only applies to
> AVR32, right?  I'm definitely not an expert here, sorry about it.
> 
> But well, if that is the case, then I guess we can either (a) treat
> AVR32 separately and make an avr32-linux-tdep.c for it, or (b) drop this
> patch entirely.

Looks like upstream GDB doesn't support AVR32 at all.  At least, I
can't find any hit for avr32 in the sources.  I looked around, and
I can't find GDB in Atmel's current toolchains (on their site).  I found
avr32-gdb-6.7.1.atmel.1.0.4.tar.gz at
http://distribute.atmel.no/tools/avr32/old/avr32gnutoolchain/v2.2.1/source/
though (note "old"), which does contain AVR32 GDB and GDBserver ports.
Guess this was never pushed upstream.  Sounds like we should drop
this patch then.

-- 
Pedro Alves

Re: [PATCH] Create avr-linux-tdep.c (add reintroduce gdbarch_gdb_signal_{to,from}_target for that target)

[Sergio Durigan Junior]

On Tuesday, August 13 2013, Pedro Alves wrote:

> Looks like upstream GDB doesn't support AVR32 at all.  At least, I
> can't find any hit for avr32 in the sources.  I looked around, and
> I can't find GDB in Atmel's current toolchains (on their site).  I found
> avr32-gdb-6.7.1.atmel.1.0.4.tar.gz at
> http://distribute.atmel.no/tools/avr32/old/avr32gnutoolchain/v2.2.1/source/
> though (note "old"), which does contain AVR32 GDB and GDBserver ports.
> Guess this was never pushed upstream.  Sounds like we should drop
> this patch then.

Thanks, Pedro.  Then consider the patch dropped.

-- 
Sergio

Re: [PATCH] Create avr-linux-tdep.c (add reintroduce gdbarch_gdb_signal_{to,from}_target for that target)

[Mike Frysinger]

[-- Attachment #1: Type: Text/Plain, Size: 418 bytes --]

On Monday 12 August 2013 17:48:57 Sergio Durigan Junior wrote:
> Maybe I am confusing things here.  I saw
> <arch/avr32/include/uapi/asm/signal.h> and I assumed one could run Linux
> on AVR, but on a further investigation I saw that it only applies to
> AVR32, right?  I'm definitely not an expert here, sorry about it.

correct, the Linux/AVR32 port is only AVR32 and not general AVR 
microcontrollers
-mike

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

Re: [PATCH] Create avr-linux-tdep.c (add reintroduce gdbarch_gdb_signal_{to,from}_target for that target)

[Sergio Durigan Junior]

On Sunday, August 25 2013, Mike Frysinger wrote:

> On Monday 12 August 2013 17:48:57 Sergio Durigan Junior wrote:
>> Maybe I am confusing things here.  I saw
>> <arch/avr32/include/uapi/asm/signal.h> and I assumed one could run Linux
>> on AVR, but on a further investigation I saw that it only applies to
>> AVR32, right?  I'm definitely not an expert here, sorry about it.
>
> correct, the Linux/AVR32 port is only AVR32 and not general AVR 
> microcontrollers

Thanks for the correction.  So it would not be wrong to create an
avr-*-linux target, right?

-- 
Sergio

Re: [PATCH] Create avr-linux-tdep.c (add reintroduce gdbarch_gdb_signal_{to,from}_target for that target)

[Mike Frysinger]

[-- Attachment #1: Type: Text/Plain, Size: 739 bytes --]

On Sunday 25 August 2013 15:42:01 Sergio Durigan Junior wrote:
> On Sunday, August 25 2013, Mike Frysinger wrote:
> > On Monday 12 August 2013 17:48:57 Sergio Durigan Junior wrote:
> >> Maybe I am confusing things here.  I saw
> >> <arch/avr32/include/uapi/asm/signal.h> and I assumed one could run Linux
> >> on AVR, but on a further investigation I saw that it only applies to
> >> AVR32, right?  I'm definitely not an expert here, sorry about it.
> > 
> > correct, the Linux/AVR32 port is only AVR32 and not general AVR
> > microcontrollers
> 
> Thanks for the correction.  So it would not be wrong to create an
> avr-*-linux target, right?

pretty sure the avr32 guys use avr32-linux-{gnu,uclibc} as their tuples
-mike

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 836 bytes --]