[Buildroot] [PATCH] m68k: add special gcc flag to avoid ICE for coldfire

[Buildroot] [PATCH] m68k: add special gcc flag to avoid ICE for coldfire

[Waldemar Brodkorb]

Fixes:
  http://autobuild.buildroot.net/results/d47fa41aa860d82471b83ac90967d3a3dacd8611
  http://autobuild.buildroot.net/results/a21de4747f40a5ce93108c8979fbc0277d040e79

Requires this as prerequisite:
http://patchwork.ozlabs.org/patch/683830/

Signed-off-by: Arnout Vandecappelle <arnout@mind.be>
Tested-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
---
 toolchain/toolchain-wrapper.mk | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/toolchain/toolchain-wrapper.mk b/toolchain/toolchain-wrapper.mk
index c7b5019..463ba42 100644
--- a/toolchain/toolchain-wrapper.mk
+++ b/toolchain/toolchain-wrapper.mk
@@ -9,6 +9,13 @@ else
 TOOLCHAIN_WRAPPER_HASH_STYLE = both
 endif
 
+TARGET_FLAGS += $(call qstrip,$(BR2_TARGET_OPTIMIZATION))
+
+# Internal error, aborting at dw2gencfi.c:214 in emit_expr_encoded
+ifeq ($(BR2_m68k_cf),y)
+TARGET_FLAGS += -fno-dwarf2-cfi-asm
+endif
+
 TOOLCHAIN_WRAPPER_ARGS = $($(PKG)_TOOLCHAIN_WRAPPER_ARGS)
 TOOLCHAIN_WRAPPER_ARGS += -DBR_SYSROOT='"$(STAGING_SUBDIR)"'
 
-- 
2.1.4

[Buildroot] [PATCH] m68k: add special gcc flag to avoid ICE for coldfire

[Arnout Vandecappelle]

On 26-11-16 09:51, Waldemar Brodkorb wrote:
> Fixes:
>   http://autobuild.buildroot.net/results/d47fa41aa860d82471b83ac90967d3a3dacd8611
>   http://autobuild.buildroot.net/results/a21de4747f40a5ce93108c8979fbc0277d040e79

 Did you test whether exception handling still works?

> 
> Requires this as prerequisite:
> http://patchwork.ozlabs.org/patch/683830/
> 
> Signed-off-by: Arnout Vandecappelle <arnout@mind.be>

 Not sure if this warrants a Sob from me, but OK.

Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

 Regards,
 Arnout

> Tested-by: Waldemar Brodkorb <wbx@openadk.org>
> Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
> ---
>  toolchain/toolchain-wrapper.mk | 7 +++++++
>  1 file changed, 7 insertions(+)
> 
> diff --git a/toolchain/toolchain-wrapper.mk b/toolchain/toolchain-wrapper.mk
> index c7b5019..463ba42 100644
> --- a/toolchain/toolchain-wrapper.mk
> +++ b/toolchain/toolchain-wrapper.mk
> @@ -9,6 +9,13 @@ else
>  TOOLCHAIN_WRAPPER_HASH_STYLE = both
>  endif
>  
> +TARGET_FLAGS += $(call qstrip,$(BR2_TARGET_OPTIMIZATION))
> +
> +# Internal error, aborting at dw2gencfi.c:214 in emit_expr_encoded
> +ifeq ($(BR2_m68k_cf),y)
> +TARGET_FLAGS += -fno-dwarf2-cfi-asm
> +endif
> +
>  TOOLCHAIN_WRAPPER_ARGS = $($(PKG)_TOOLCHAIN_WRAPPER_ARGS)
>  TOOLCHAIN_WRAPPER_ARGS += -DBR_SYSROOT='"$(STAGING_SUBDIR)"'
>  
> 

-- 
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF

[Buildroot] [PATCH] m68k: add special gcc flag to avoid ICE for coldfire

[Thomas Petazzoni]

Hello,

On Sat, 26 Nov 2016 09:51:44 +0100, Waldemar Brodkorb wrote:
> Fixes:
>   http://autobuild.buildroot.net/results/d47fa41aa860d82471b83ac90967d3a3dacd8611
>   http://autobuild.buildroot.net/results/a21de4747f40a5ce93108c8979fbc0277d040e79
> 
> Requires this as prerequisite:
> http://patchwork.ozlabs.org/patch/683830/

Hum, thanks but this patch you're pointing to is really material for
next at this point, so I cannot apply your patch, which should be
applied on master in order to fix the autobuilder issues.

Thanks,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com

[Buildroot] [PATCH] m68k: add special gcc flag to avoid ICE for coldfire

[Arnout Vandecappelle]

On 26-11-16 14:52, Thomas Petazzoni wrote:
> Hello,
> 
> On Sat, 26 Nov 2016 09:51:44 +0100, Waldemar Brodkorb wrote:
>> Fixes:
>>   http://autobuild.buildroot.net/results/d47fa41aa860d82471b83ac90967d3a3dacd8611
>>   http://autobuild.buildroot.net/results/a21de4747f40a5ce93108c8979fbc0277d040e79
>>
>> Requires this as prerequisite:
>> http://patchwork.ozlabs.org/patch/683830/
> 
> Hum, thanks but this patch you're pointing to is really material for
> next at this point, so I cannot apply your patch, which should be
> applied on master in order to fix the autobuilder issues.

 Begging to differ: doing the same without patch 683830 would be a much bigger
patch (adding yet another -D option to the wrapper, updating the wrapper
itself). But OK, it's getting a bit later to still take _any_ risk, so probably
the bigger patch is better.

 Regards,
 Arnout

> 
> Thanks,
> 
> Thomas
> 

-- 
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF

[Buildroot] [PATCH] m68k: add special gcc flag to avoid ICE for coldfire

[Thomas Petazzoni]

Hello,

On Sat, 26 Nov 2016 14:59:34 +0100, Arnout Vandecappelle wrote:

> > Hum, thanks but this patch you're pointing to is really material for
> > next at this point, so I cannot apply your patch, which should be
> > applied on master in order to fix the autobuilder issues.  
> 
>  Begging to differ: doing the same without patch 683830 would be a much bigger
> patch (adding yet another -D option to the wrapper, updating the wrapper
> itself). But OK, it's getting a bit later to still take _any_ risk, so probably
> the bigger patch is better.

That's my point: that late in the cycle, I don't want to touch too
many things around the toolchain wrapper. So adding one more CFLAGS or
LDFLAGS specific to m68k is OK (worst thing is a regression on m68k,
which probably isn't very widely used). But a change that touches the
toolchain wrapper, and can potentially break major architectures, I'm
not a big fan.

If nobody provides the bigger patch, I think I'm going to live with the
m68k autobuilder failures until the release, and merge afterwards.

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com

[Buildroot] [PATCH] m68k: add special gcc flag to avoid ICE for coldfire

[Waldemar Brodkorb]

Hi Arnout,
Arnout Vandecappelle wrote,

> On 26-11-16 09:51, Waldemar Brodkorb wrote:
> > Fixes:
> >   http://autobuild.buildroot.net/results/d47fa41aa860d82471b83ac90967d3a3dacd8611
> >   http://autobuild.buildroot.net/results/a21de4747f40a5ce93108c8979fbc0277d040e79
> 
>  Did you test whether exception handling still works?

Tested with attached C++ app.
Before and after the patches.
 
> > Requires this as prerequisite:
> > http://patchwork.ozlabs.org/patch/683830/
> > 
> > Signed-off-by: Arnout Vandecappelle <arnout@mind.be>
> 
>  Not sure if this warrants a Sob from me, but OK.
> 
> Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

Will you provide the bigger patch Thomas wants to have?

best regards
 Waldemar
-------------- next part --------------
A non-text attachment was scrubbed...
Name: foo.cc
Type: text/x-c++src
Size: 2835 bytes
Desc: not available
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20161127/64501825/attachment.cc>

[Buildroot] [PATCH] m68k: add special gcc flag to avoid ICE for coldfire

[Thomas Petazzoni]

Hello,

On Sat, 26 Nov 2016 14:59:34 +0100, Arnout Vandecappelle wrote:

>  Begging to differ: doing the same without patch 683830 would be a much bigger
> patch (adding yet another -D option to the wrapper, updating the wrapper
> itself). But OK, it's getting a bit later to still take _any_ risk, so probably
> the bigger patch is better.

Have you had the chance to work on your toolchain wrapper patches, at
least enough to be able to merge this m68k change?

Thanks!

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com

[Buildroot] [PATCH] m68k: add special gcc flag to avoid ICE for coldfire

[Arnout Vandecappelle]

On 11-12-16 15:51, Thomas Petazzoni wrote:
> Hello,
> 
> On Sat, 26 Nov 2016 14:59:34 +0100, Arnout Vandecappelle wrote:
> 
>>  Begging to differ: doing the same without patch 683830 would be a much bigger
>> patch (adding yet another -D option to the wrapper, updating the wrapper
>> itself). But OK, it's getting a bit later to still take _any_ risk, so probably
>> the bigger patch is better.
> 
> Have you had the chance to work on your toolchain wrapper patches, at
> least enough to be able to merge this m68k change?

 Not yet, and it doesn't look like I'll have much time this year...

 Regards,
 Arnout

-- 
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF

[Buildroot] [PATCH] m68k: add special gcc flag to avoid ICE for coldfire

[Thomas Petazzoni]

Hello,

On Sat, 26 Nov 2016 09:51:44 +0100, Waldemar Brodkorb wrote:
> Fixes:
>   http://autobuild.buildroot.net/results/d47fa41aa860d82471b83ac90967d3a3dacd8611
>   http://autobuild.buildroot.net/results/a21de4747f40a5ce93108c8979fbc0277d040e79
> 
> Requires this as prerequisite:
> http://patchwork.ozlabs.org/patch/683830/
> 
> Signed-off-by: Arnout Vandecappelle <arnout@mind.be>
> Tested-by: Waldemar Brodkorb <wbx@openadk.org>
> Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>

We've added -fno-dwarf2-cfi-asm to 4 packages so far, and it seems to
have fixed the problem. So I would for now prefer to keep this as it is
today, and not use the wrapper to pass this flag to all packages.

I'll mark your patch as Rejected in patchwork, since it anyway cannot
be applied because Arnout patches have not been merged.

Let me know if you disagree, of course.

Thanks!

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com