[PATCH 1/2] Merge config/ changes from GCC, to enable DFP on AArch64

[PATCH 1/2] Merge config/ changes from GCC, to enable DFP on AArch64

[Christophe Lyon]

Hi,

I have just added support in GCC for Decimal Floating Point (DFP) on AArch64.
These two patches enable it in GDB:

1- Merge the necessary update to dfp.m4 from GCC, so that DFP is
   enabled at configure time.

2- Add DFP support in GDB when calling or returning from a function
   with DFP arguments or return value.

2022-04-28  Christophe Lyon  <christophe.lyon@arm.com>

	config/
	* dfp.m4 (enable_decimal_float): Enable BID for AArch64.

	libdecnumber/
	* configure: Regenerate.
---
 config/dfp.m4          | 21 ++++++++++++++++++++-
 libdecnumber/configure |  3 ++-
 2 files changed, 22 insertions(+), 2 deletions(-)

diff --git a/config/dfp.m4 b/config/dfp.m4
index a137ddebf8c..d1d151b2da7 100644
--- a/config/dfp.m4
+++ b/config/dfp.m4
@@ -1,3 +1,21 @@
+# Copyright (C) 2010-2022 Free Software Foundation, Inc.
+
+# This file is part of GCC.
+
+# GCC 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, or (at your option) any
+# later version.
+
+# GCC 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 GCC; see the file COPYING3.  If not see
+# <http://www.gnu.org/licenses/>.
+
 dnl @synopsis GCC_AC_ENABLE_DECIMAL_FLOAT([target triplet])
 dnl
 dnl Enable C extension for decimal float if target supports it.
@@ -20,6 +38,7 @@ Valid choices are 'yes', 'bid', 'dpd', and 'no'.]) ;;
 ],
 [
   case $1 in
+    aarch64* | \
     powerpc*-*-linux* | i?86*-*-linux* | x86_64*-*-linux* | s390*-*-linux* | \
     i?86*-*-elfiamcu | i?86*-*-gnu* | x86_64*-*-gnu* | \
     i?86*-*-mingw* | x86_64*-*-mingw* | \
@@ -37,7 +56,7 @@ Valid choices are 'yes', 'bid', 'dpd', and 'no'.]) ;;
 case x$enable_decimal_float in
   xyes)
     case $1 in
-      i?86*-*-* | x86_64*-*-*)
+      aarch64* | i?86*-*-* | x86_64*-*-*)
 	enable_decimal_float=bid
 	;;
       *)
diff --git a/libdecnumber/configure b/libdecnumber/configure
index 14c103a0718..87f3bd5e457 100755
--- a/libdecnumber/configure
+++ b/libdecnumber/configure
@@ -4761,6 +4761,7 @@ Valid choices are 'yes', 'bid', 'dpd', and 'no'." "$LINENO" 5 ;;
 else
 
   case $target in
+    aarch64* | \
     powerpc*-*-linux* | i?86*-*-linux* | x86_64*-*-linux* | s390*-*-linux* | \
     i?86*-*-elfiamcu | i?86*-*-gnu* | x86_64*-*-gnu* | \
     i?86*-*-mingw* | x86_64*-*-mingw* | \
@@ -4781,7 +4782,7 @@ fi
 case x$enable_decimal_float in
   xyes)
     case $target in
-      i?86*-*-* | x86_64*-*-*)
+      aarch64* | i?86*-*-* | x86_64*-*-*)
 	enable_decimal_float=bid
 	;;
       *)
-- 
2.25.1

[PATCH 2/2] AArch64: add support for DFP (Decimal Floating point)

[Christophe Lyon]

This small patch adds support for TYPE_CODE_DECFLOAT in
aapcs_is_vfp_call_or_return_candidate_1 and pass_in_v_vfp_candidate,
so that GDB for AArch64 knows how to pass DFP parameters and how to
read DFP results when calling a function.

Tested on aarch64-linux-gnu, with a GCC with DFP support in the PATH,
all of GDB's DFP tests pass.
---
 gdb/aarch64-tdep.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/gdb/aarch64-tdep.c b/gdb/aarch64-tdep.c
index e35c083454d..67a3f96e1a7 100644
--- a/gdb/aarch64-tdep.c
+++ b/gdb/aarch64-tdep.c
@@ -1470,6 +1470,7 @@ aapcs_is_vfp_call_or_return_candidate_1 (struct type *type,
   switch (type->code ())
     {
     case TYPE_CODE_FLT:
+    case TYPE_CODE_DECFLOAT:
       if (TYPE_LENGTH (type) > 16)
 	return -1;
 
@@ -1780,6 +1781,7 @@ pass_in_v_vfp_candidate (struct gdbarch *gdbarch, struct regcache *regcache,
   switch (arg_type->code ())
     {
     case TYPE_CODE_FLT:
+    case TYPE_CODE_DECFLOAT:
       return pass_in_v (gdbarch, regcache, info, TYPE_LENGTH (arg_type),
 			value_contents (arg).data ());
       break;
-- 
2.25.1

Re: [PATCH 1/2] Merge config/ changes from GCC, to enable DFP on AArch64

[Tom Tromey]

>>>>> "Christophe" == Christophe Lyon via Gdb-patches <gdb-patches@sourceware.org> writes:

Christophe> I have just added support in GCC for Decimal Floating Point (DFP) on AArch64.
Christophe> These two patches enable it in GDB:

If this went in to GCC, then it's fine here as well.  Thank you.

Tom

Re: [PATCH 2/2] AArch64: add support for DFP (Decimal Floating point)

[Christophe Lyon]

On 5/20/22 11:06, Christophe Lyon wrote:
> This small patch adds support for TYPE_CODE_DECFLOAT in
> aapcs_is_vfp_call_or_return_candidate_1 and pass_in_v_vfp_candidate,
> so that GDB for AArch64 knows how to pass DFP parameters and how to
> read DFP results when calling a function.
> 
> Tested on aarch64-linux-gnu, with a GCC with DFP support in the PATH,
> all of GDB's DFP tests pass.

More precisely, patch 1/2 has a side-effect of enabling the GDB DFP 
tests, some of which fail without this patch 2/2:
gdb.base/dfp-test.exp: Call function with many _Decimal128 arguments.
gdb.base/dfp-test.exp: Call function with many _Decimal32 arguments.
gdb.base/dfp-test.exp: Call function with many _Decimal64 arguments.
gdb.base/dfp-test.exp: Call function with many mixed decimal float 
arguments.
gdb.base/dfp-test.exp: Call function with mixed decimal float arguments 
TEST.
gdb.base/dfp-test.exp: Call function with mixed decimal float arguments.
gdb.base/dfp-test.exp: backtrace function with correct _Decimal128 
arguments.
gdb.base/dfp-test.exp: backtrace function with correct _Decimal32 arguments.
gdb.base/dfp-test.exp: backtrace function with correct _Decimal64 arguments.

OK?

Christophe


> ---
>   gdb/aarch64-tdep.c | 2 ++
>   1 file changed, 2 insertions(+)
> 
> diff --git a/gdb/aarch64-tdep.c b/gdb/aarch64-tdep.c
> index e35c083454d..67a3f96e1a7 100644
> --- a/gdb/aarch64-tdep.c
> +++ b/gdb/aarch64-tdep.c
> @@ -1470,6 +1470,7 @@ aapcs_is_vfp_call_or_return_candidate_1 (struct type *type,
>     switch (type->code ())
>       {
>       case TYPE_CODE_FLT:
> +    case TYPE_CODE_DECFLOAT:
>         if (TYPE_LENGTH (type) > 16)
>   	return -1;
>   
> @@ -1780,6 +1781,7 @@ pass_in_v_vfp_candidate (struct gdbarch *gdbarch, struct regcache *regcache,
>     switch (arg_type->code ())
>       {
>       case TYPE_CODE_FLT:
> +    case TYPE_CODE_DECFLOAT:
>         return pass_in_v (gdbarch, regcache, info, TYPE_LENGTH (arg_type),
>   			value_contents (arg).data ());
>         break;

Re: [PATCH 2/2] AArch64: add support for DFP (Decimal Floating point)

[Luis Machado]

On 5/24/22 08:55, Christophe Lyon wrote:
> 
> 
> On 5/20/22 11:06, Christophe Lyon wrote:
>> This small patch adds support for TYPE_CODE_DECFLOAT in
>> aapcs_is_vfp_call_or_return_candidate_1 and pass_in_v_vfp_candidate,
>> so that GDB for AArch64 knows how to pass DFP parameters and how to
>> read DFP results when calling a function.
>>
>> Tested on aarch64-linux-gnu, with a GCC with DFP support in the PATH,
>> all of GDB's DFP tests pass.
> 
> More precisely, patch 1/2 has a side-effect of enabling the GDB DFP tests, some of which fail without this patch 2/2:
> gdb.base/dfp-test.exp: Call function with many _Decimal128 arguments.
> gdb.base/dfp-test.exp: Call function with many _Decimal32 arguments.
> gdb.base/dfp-test.exp: Call function with many _Decimal64 arguments.
> gdb.base/dfp-test.exp: Call function with many mixed decimal float arguments.
> gdb.base/dfp-test.exp: Call function with mixed decimal float arguments TEST.
> gdb.base/dfp-test.exp: Call function with mixed decimal float arguments.
> gdb.base/dfp-test.exp: backtrace function with correct _Decimal128 arguments.
> gdb.base/dfp-test.exp: backtrace function with correct _Decimal32 arguments.
> gdb.base/dfp-test.exp: backtrace function with correct _Decimal64 arguments.
> 
> OK?

This is OK. Thanks!