[PATCH, ARM] Don't pull in unwinder for 64-bit division routines

[PATCH, ARM] Don't pull in unwinder for 64-bit division routines

[Julian Brown]

[-- Attachment #1: Type: text/plain, Size: 2166 bytes --]

Hi,

This (old!) patch avoids uselessly pulling in the unwinder for 64-bit
division routines. I last posted it here:

http://gcc.gnu.org/ml/gcc-patches/2009-10/msg01618.html

Other people have noticed the same issue, e.g.:

http://gcc.gnu.org/ml/gcc-help/2011-03/msg00187.html

Note that we take special care already for the case where
division-by-zero throws an exception: we tailcall the division-by-zero
handler at the head of the 64-bit division helper routines, so never
create a stack frame which needs to be unwound. On ARM, division by
zero can trapped by defining magic __aeabi_[il]div0 functions. See
here:

http://gcc.gnu.org/ml/gcc-patches/2009-10/msg01661.html

Unfortunately, despite that, throwing an exception from the
division-by-zero handler functions doesn't appear to work at present --
either with or without the current patch. I'm not intending to tackle
that now (see attached program -- which exits with "terminate called
after throwing an instance of 'int'" for me). I'm no C++ expert, so I
might have done something dumb :-).

Anyway: this revised version of the patch removes the strange libgcc
Makefile-fragment changes, the equivalent of which have since been
incorporated into mainline GCC now anyway, so the patch is somewhat
more straightforward than it was previously.

The patch allows a trivial program using 64-bit division (on bare metal)
to shrink quite considerably:

$ arm-none-eabi-size div64
   text    data     bss     dec     hex filename
  13120    2404     276   15800    3db8 div64

$ arm-none-eabi-size div64-unpatched 
   text    data     bss     dec     hex filename
  20064    2404     276   22744    58d8 div64-unpatched

OK to apply?

Thanks,

Julian

ChangeLog

    libgcc/
    * config.host (arm*-*-linux-*eabi, arm*-*-uclinux*eabi)
    (arm*-*-eabi*, arm*-*-symbianelf*, arm*-*-rtemseabi*): Add
    arm/t-divmod-ef to tmake_file.
    * Makefile.in (LIB2_DIVMOD_EXCEPTION_FLAGS): Default to -fexceptions
    -fnon-call-exceptions if not defined.
    ($(lib2-divmod-o), $(lib2-divmod-s-o)): Use above.
    * config/arm/t-divmod-ef: New file.

    gcc/testsuite/
    * gcc.target/arm/div64-unwinding.c: New test.

[-- Attachment #2: div64.cc --]
[-- Type: text/x-c++src, Size: 396 bytes --]

#include <cstdio>

long long a = 100, b = 0;

extern "C" long long
__aeabi_ldiv0 (long long retval)
{
  throw 0;
}

int bad_division (void)
{
  return a / b;
}

int main(int argc, char *argv[])
{
  try
    {
      bad_division ();
    }
  catch (int e)
    {
      fprintf (stderr, "caught division by zero\n");
      return 0;
    }
  fprintf (stderr, "didn't catch exception\n");
  return 1;
}

[-- Attachment #3: no-unwinder-for-long-division-fsf-3.diff --]
[-- Type: text/x-patch, Size: 3574 bytes --]

Index: libgcc/config.host
===================================================================
--- libgcc/config.host	(revision 189656)
+++ libgcc/config.host	(working copy)
@@ -317,7 +317,7 @@ arm*-*-linux*)			# ARM GNU/Linux with EL
 	tmake_file="${tmake_file} arm/t-arm t-fixedpoint-gnu-prefix"
 	case ${host} in
 	arm*-*-linux-*eabi)
-	  tmake_file="${tmake_file} arm/t-elf arm/t-bpabi arm/t-linux-eabi t-slibgcc-libgcc"
+	  tmake_file="${tmake_file} arm/t-elf arm/t-bpabi arm/t-linux-eabi arm/t-divmod-ef t-slibgcc-libgcc"
 	  tm_file="$tm_file arm/bpabi-lib.h"
 	  unwind_header=config/arm/unwind-arm.h
 	  ;;
@@ -331,7 +331,7 @@ arm*-*-uclinux*)		# ARM ucLinux
 	tmake_file="${tmake_file} t-fixedpoint-gnu-prefix"
 	case ${host} in
 	arm*-*-uclinux*eabi)
-	  tmake_file="${tmake_file} arm/t-bpabi"
+	  tmake_file="${tmake_file} arm/t-bpabi arm/t-divmod-ef"
 	  tm_file="$tm_file arm/bpabi-lib.h"
 	  unwind_header=config/arm/unwind-arm.h
 	  ;;
@@ -344,7 +344,7 @@ arm*-*-ecos-elf)
 	extra_parts="$extra_parts crti.o crtn.o"
 	;;
 arm*-*-eabi* | arm*-*-symbianelf* | arm*-*-rtemseabi*)
-	tmake_file="${tmake_file} arm/t-arm arm/t-elf t-fixedpoint-gnu-prefix"
+	tmake_file="${tmake_file} arm/t-arm arm/t-elf arm/t-divmod-ef t-fixedpoint-gnu-prefix"
 	tm_file="$tm_file arm/bpabi-lib.h"
 	case ${host} in
 	arm*-*-eabi* | arm*-*-rtemseabi*)
Index: libgcc/Makefile.in
===================================================================
--- libgcc/Makefile.in	(revision 189656)
+++ libgcc/Makefile.in	(working copy)
@@ -497,18 +497,24 @@ libgcc-s-objects += $(patsubst %,%_s$(ob
 endif
 endif
 
+ifeq ($(LIB2_DIVMOD_EXCEPTION_FLAGS),)
+# Provide default flags for compiling divmod functions, if they haven't been
+# set already by a target-specific Makefile fragment.
+LIB2_DIVMOD_EXCEPTION_FLAGS := -fexceptions -fnon-call-exceptions
+endif
+
 # Build LIB2_DIVMOD_FUNCS.
 lib2-divmod-o = $(patsubst %,%$(objext),$(LIB2_DIVMOD_FUNCS))
 $(lib2-divmod-o): %$(objext): $(srcdir)/libgcc2.c
 	$(gcc_compile) -DL$* -c $< \
-	  -fexceptions -fnon-call-exceptions $(vis_hide)
+	  $(LIB2_DIVMOD_EXCEPTION_FLAGS) $(vis_hide)
 libgcc-objects += $(lib2-divmod-o)
 
 ifeq ($(enable_shared),yes)
 lib2-divmod-s-o = $(patsubst %,%_s$(objext),$(LIB2_DIVMOD_FUNCS))
 $(lib2-divmod-s-o): %_s$(objext): $(srcdir)/libgcc2.c
 	$(gcc_s_compile) -DL$* -c $< \
-	  -fexceptions -fnon-call-exceptions
+	  $(LIB2_DIVMOD_EXCEPTION_FLAGS)
 libgcc-s-objects += $(lib2-divmod-s-o)
 endif
 
Index: libgcc/config/arm/t-divmod-ef
===================================================================
--- libgcc/config/arm/t-divmod-ef	(revision 0)
+++ libgcc/config/arm/t-divmod-ef	(revision 0)
@@ -0,0 +1,4 @@
+# On ARM, specifying -fnon-call-exceptions will needlessly pull in
+# the unwinder in simple programs which use 64-bit division.  Omitting
+# the option is safe.
+LIB2_DIVMOD_EXCEPTION_FLAGS := -fexceptions
Index: gcc/testsuite/gcc.target/arm/div64-unwinding.c
===================================================================
--- gcc/testsuite/gcc.target/arm/div64-unwinding.c	(revision 0)
+++ gcc/testsuite/gcc.target/arm/div64-unwinding.c	(revision 0)
@@ -0,0 +1,24 @@
+/* Performing a 64-bit division should not pull in the unwinder.  */
+
+/* { dg-do run } */
+/* { dg-options "-O0" } */
+
+#include <stdlib.h>
+
+long long
+foo (long long c, long long d)
+{
+  return c/d;
+}
+
+long long x = 0;
+long long y = 1;
+
+extern int (*_Unwind_RaiseException) (void *) __attribute__((weak));
+
+int main(void)
+{
+  if (&_Unwind_RaiseException != NULL)
+    abort ();;
+  return foo (x, y);
+}

Re: [PATCH, ARM] Don't pull in unwinder for 64-bit division routines

[Julian Brown]

[-- Attachment #1: Type: text/plain, Size: 883 bytes --]

On Fri, 20 Jul 2012 11:15:27 +0100
Julian Brown <julian@codesourcery.com> wrote:

> Anyway: this revised version of the patch removes the strange libgcc
> Makefile-fragment changes, the equivalent of which have since been
> incorporated into mainline GCC now anyway, so the patch is somewhat
> more straightforward than it was previously.

Joey Ye contacted me offlist and suggested that the t-divmod-ef
fragment might be better integrated into t-bpabi instead. Doing that
makes the patch somewhat smaller/cleaner.

Minimally re-tested, looks OK.

Julian

ChangeLog

    libgcc/
    * Makefile.in (LIB2_DIVMOD_EXCEPTION_FLAGS): Default to
    -fexceptions -fnon-call-exceptions if not defined.
    ($(lib2-divmod-o), $(lib2-divmod-s-o)): Use above.
    * config/arm/t-bpabi (LIB2_DIVMOD_EXCEPTION_FLAGS): Define.

    gcc/testsuite/
    * gcc.target/arm/div64-unwinding.c: New test.

[-- Attachment #2: no-unwinder-for-long-division-fsf-4.diff --]
[-- Type: text/x-patch, Size: 2352 bytes --]

Index: libgcc/Makefile.in
===================================================================
--- libgcc/Makefile.in	(revision 189807)
+++ libgcc/Makefile.in	(working copy)
@@ -497,18 +497,24 @@ libgcc-s-objects += $(patsubst %,%_s$(ob
 endif
 endif
 
+ifeq ($(LIB2_DIVMOD_EXCEPTION_FLAGS),)
+# Provide default flags for compiling divmod functions, if they haven't been
+# set already by a target-specific Makefile fragment.
+LIB2_DIVMOD_EXCEPTION_FLAGS := -fexceptions -fnon-call-exceptions
+endif
+
 # Build LIB2_DIVMOD_FUNCS.
 lib2-divmod-o = $(patsubst %,%$(objext),$(LIB2_DIVMOD_FUNCS))
 $(lib2-divmod-o): %$(objext): $(srcdir)/libgcc2.c
 	$(gcc_compile) -DL$* -c $< \
-	  -fexceptions -fnon-call-exceptions $(vis_hide)
+	  $(LIB2_DIVMOD_EXCEPTION_FLAGS) $(vis_hide)
 libgcc-objects += $(lib2-divmod-o)
 
 ifeq ($(enable_shared),yes)
 lib2-divmod-s-o = $(patsubst %,%_s$(objext),$(LIB2_DIVMOD_FUNCS))
 $(lib2-divmod-s-o): %_s$(objext): $(srcdir)/libgcc2.c
 	$(gcc_s_compile) -DL$* -c $< \
-	  -fexceptions -fnon-call-exceptions
+	  $(LIB2_DIVMOD_EXCEPTION_FLAGS)
 libgcc-s-objects += $(lib2-divmod-s-o)
 endif
 
Index: libgcc/config/arm/t-bpabi
===================================================================
--- libgcc/config/arm/t-bpabi	(revision 189807)
+++ libgcc/config/arm/t-bpabi	(working copy)
@@ -13,3 +13,8 @@ LIB2ADDEH = $(srcdir)/config/arm/unwind-
 
 # Add the BPABI names.
 SHLIB_MAPFILES += $(srcdir)/config/arm/libgcc-bpabi.ver
+
+# On ARM, specifying -fnon-call-exceptions will needlessly pull in
+# the unwinder in simple programs which use 64-bit division.  Omitting
+# the option is safe.
+LIB2_DIVMOD_EXCEPTION_FLAGS := -fexceptions
Index: gcc/testsuite/gcc.target/arm/div64-unwinding.c
===================================================================
--- gcc/testsuite/gcc.target/arm/div64-unwinding.c	(revision 0)
+++ gcc/testsuite/gcc.target/arm/div64-unwinding.c	(revision 0)
@@ -0,0 +1,24 @@
+/* Performing a 64-bit division should not pull in the unwinder.  */
+
+/* { dg-do run } */
+/* { dg-options "-O0" } */
+
+#include <stdlib.h>
+
+long long
+foo (long long c, long long d)
+{
+  return c/d;
+}
+
+long long x = 0;
+long long y = 1;
+
+extern int (*_Unwind_RaiseException) (void *) __attribute__((weak));
+
+int main(void)
+{
+  if (&_Unwind_RaiseException != NULL)
+    abort ();;
+  return foo (x, y);
+}

Re: [PATCH, ARM] Don't pull in unwinder for 64-bit division routines

[Sebastian Huber]

Hello,

with this move to t-bpabi other targets like RTEMS profit also from this 
change.  This is very good since the unwinder pull-in for 64-bit divisions was 
pretty bad for small Cortex-M3 systems with internal flash only.

-- 
Sebastian Huber, embedded brains GmbH

Address : Obere Lagerstr. 30, D-82178 Puchheim, Germany
Phone   : +49 89 18 90 80 79-6
Fax     : +49 89 18 90 80 79-9
E-Mail  : sebastian.huber@embedded-brains.de
PGP     : Public key available on request.

Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.

Re: [PATCH, ARM] Don't pull in unwinder for 64-bit division routines

[Ramana Radhakrishnan]

On 07/24/12 13:27, Julian Brown wrote:
> On Fri, 20 Jul 2012 11:15:27 +0100
> Julian Brown <julian@codesourcery.com> wrote:
>
>> Anyway: this revised version of the patch removes the strange libgcc
>> Makefile-fragment changes, the equivalent of which have since been
>> incorporated into mainline GCC now anyway, so the patch is somewhat
>> more straightforward than it was previously.
>
> Joey Ye contacted me offlist and suggested that the t-divmod-ef
> fragment might be better integrated into t-bpabi instead. Doing that
> makes the patch somewhat smaller/cleaner.
>
> Minimally re-tested, looks OK.

The original submission makes no mention of testing ? The ARM specific 
portions look OK to me modulo no regressions.

Ian, can also take a quick look.

regards,
Ramana

Re: [PATCH, ARM] Don't pull in unwinder for 64-bit division routines

[Julian Brown]

On Thu, 16 Aug 2012 19:56:52 +0100
Ramana Radhakrishnan <ramrad01@arm.com> wrote:

> On 07/24/12 13:27, Julian Brown wrote:
> > On Fri, 20 Jul 2012 11:15:27 +0100
> > Julian Brown <julian@codesourcery.com> wrote:
> >
> >> Anyway: this revised version of the patch removes the strange
> >> libgcc Makefile-fragment changes, the equivalent of which have
> >> since been incorporated into mainline GCC now anyway, so the patch
> >> is somewhat more straightforward than it was previously.
> >
> > Joey Ye contacted me offlist and suggested that the t-divmod-ef
> > fragment might be better integrated into t-bpabi instead. Doing that
> > makes the patch somewhat smaller/cleaner.
> >
> > Minimally re-tested, looks OK.
> 
> The original submission makes no mention of testing ? The ARM
> specific portions look OK to me modulo no regressions.

Thanks -- I'm sure I did test the patch, but just omitted to mention
that fact in the mail :-O. We've also been carrying a version of this
patch in our local source base for many years now.

> Ian, can also take a quick look.

Cheers,

Julian

Re: [PATCH, ARM] Don't pull in unwinder for 64-bit division routines

[Ian Lance Taylor]

On Thu, Aug 16, 2012 at 11:56 AM, Ramana Radhakrishnan <ramrad01@arm.com> wrote:
> On 07/24/12 13:27, Julian Brown wrote:
>>
>> On Fri, 20 Jul 2012 11:15:27 +0100
>> Julian Brown <julian@codesourcery.com> wrote:
>>
>>> Anyway: this revised version of the patch removes the strange libgcc
>>> Makefile-fragment changes, the equivalent of which have since been
>>> incorporated into mainline GCC now anyway, so the patch is somewhat
>>> more straightforward than it was previously.
>>
>>
>> Joey Ye contacted me offlist and suggested that the t-divmod-ef
>> fragment might be better integrated into t-bpabi instead. Doing that
>> makes the patch somewhat smaller/cleaner.
>>
>> Minimally re-tested, looks OK.
>
>
> The original submission makes no mention of testing ? The ARM specific
> portions look OK to me modulo no regressions.
>
> Ian, can also take a quick look.

Looks fine to me.

Ian

Re: [PATCH, ARM] Don't pull in unwinder for 64-bit division routines

[Ye Joey]

On Fri, Aug 17, 2012 at 9:13 AM, Ian Lance Taylor <iant@google.com> wrote:
>
> Looks fine to me.
>
> Ian
Will backport to arm/embedded-4_7-branch. No sure if appropriate for
4.7 branch since it is not a stability problem.

- Joey

Re: [PATCH, ARM] Don't pull in unwinder for 64-bit division routines

[Michael Hope]

On 17 August 2012 07:29, Julian Brown <julian@codesourcery.com> wrote:
> On Thu, 16 Aug 2012 19:56:52 +0100
> Ramana Radhakrishnan <ramrad01@arm.com> wrote:
>
>> On 07/24/12 13:27, Julian Brown wrote:
>> > On Fri, 20 Jul 2012 11:15:27 +0100
>> > Julian Brown <julian@codesourcery.com> wrote:
>> >
>> >> Anyway: this revised version of the patch removes the strange
>> >> libgcc Makefile-fragment changes, the equivalent of which have
>> >> since been incorporated into mainline GCC now anyway, so the patch
>> >> is somewhat more straightforward than it was previously.
>> >
>> > Joey Ye contacted me offlist and suggested that the t-divmod-ef
>> > fragment might be better integrated into t-bpabi instead. Doing that
>> > makes the patch somewhat smaller/cleaner.
>> >
>> > Minimally re-tested, looks OK.
>>
>> The original submission makes no mention of testing ? The ARM
>> specific portions look OK to me modulo no regressions.
>
> Thanks -- I'm sure I did test the patch, but just omitted to mention
> that fact in the mail :-O. We've also been carrying a version of this
> patch in our local source base for many years now.

Hi Julian.  The test case fails on arm-linux-gnueabi:
 http://gcc.gnu.org/ml/gcc-testresults/2012-08/msg02100.html

FAIL: gcc.target/arm/div64-unwinding.c execution test

The test aborts as &_Unwind_RaiseException is not null.  _divdi3.o
itself looks fine and no longer pulls in the unwinder so I assume
something else in the environment is.  I've put the binaries up at
http://people.linaro.org/~michaelh/incoming/div64-unwinding.exe and
http://people.linaro.org/~michaelh/incoming/_divdi3.o if that helps.

-- Michael