Hi Geert!
On 6/19/20 1:19 PM, Geert Uytterhoeven wrote:
> So now you have 5 bytes of padding if PERL_DEBUG_READONLY_OPS,
> while I guess 1 would be sufficient?
>
> What about
>
> # ifdef PERL_DEBUG_READONLY_OPS
> bool opslab_readonly;
> U8 opslab_padding;
>
Hi Adrian,
On Fri, Jun 19, 2020 at 12:08 PM John Paul Adrian Glaubitz
wrote:
> This works, both with PERL_DEBUG_READONLY_OPS and without:
>
> diff --git a/op.h b/op.h
> index fc21f03cda..fc7e73fef4 100644
> --- a/op.h
> +++ b/op.h
> @@ -713,7 +713,9 @@ struct opslab {
>
Hi!
On 6/19/20 12:08 PM, John Paul Adrian Glaubitz wrote:
> This works, both with PERL_DEBUG_READONLY_OPS and without:
Updated patch attached.
Adrian
--
.''`. John Paul Adrian Glaubitz
: :' : Debian Developer - glaub...@debian.org
`. `' Freie Universitaet Berlin - glaub...@physik.fu-berli
Hi!
This works, both with PERL_DEBUG_READONLY_OPS and without:
diff --git a/op.h b/op.h
index fc21f03cda..fc7e73fef4 100644
--- a/op.h
+++ b/op.h
@@ -713,7 +713,9 @@ struct opslab {
units) */
# ifdef PERL_DEBUG_READONLY_OPS
bool opslab_reado
Hi Laurent!
On 6/19/20 11:33 AM, Laurent Vivier wrote:
>
> Can't you move "opslab_slots" before "# ifdef PERL_DEBUG_READONLY_OPS"?
I just tried this variant, it also crashes:
diff --git a/op.h b/op.h
index fc21f03cda..69f74843bf 100644
--- a/op.h
+++ b/op.h
@@ -711,10 +711,12 @@ struct opslab {
Le 19/06/2020 à 11:08, John Paul Adrian Glaubitz a écrit :
> On 6/19/20 11:03 AM, Geert Uytterhoeven wrote:
>> Either invert the order of the two paddings, or replace them by a single one:
>>
>> U8 _padding[3];
> I just tried this variant and it didn't help:
>
> diff --git a/op.h b/op.h
> ind
On 6/19/20 11:03 AM, Geert Uytterhoeven wrote:
> Either invert the order of the two paddings, or replace them by a single one:
>
> U8 _padding[3];
I just tried this variant and it didn't help:
diff --git a/op.h b/op.h
index fc21f03cda..703be6b3f2 100644
--- a/op.h
+++ b/op.h
@@ -713,6 +713,7
Hi Adrian,
On Fri, Jun 19, 2020 at 10:52 AM John Paul Adrian Glaubitz
wrote:
> On 6/19/20 10:47 AM, John Paul Adrian Glaubitz wrote:
> > It might be set during stage1, but not stage2.
> >
> > I will have to look at it again.
>
> I just tried this change and it doesn't fix the problem:
That's act
On 6/19/20 10:47 AM, John Paul Adrian Glaubitz wrote:
> It might be set during stage1, but not stage2.
>
> I will have to look at it again.
I just tried this change and it doesn't fix the problem:
diff --git a/op.h b/op.h
index fc21f03cda..ca3102d7d2 100644
--- a/op.h
+++ b/op.h
@@ -713,6 +713,8
On 6/19/20 10:42 AM, Geert Uytterhoeven wrote:
> So release builds fail, too?
Well, the build fails during the stage1 build.
> And they don't set PERL_DEBUG_READONLY_OPS?
>
> If PERL_DEBUG_READONLY_OPS is not set, I see no reason why
> the padding would be needed.
It might be set during stage1,
Hi Adrian,
On Fri, Jun 19, 2020 at 10:20 AM John Paul Adrian Glaubitz
wrote:
> > On Jun 19, 2020, at 10:02 AM, Geert Uytterhoeven
> > wrote:
> >> On Thu, Jun 18, 2020 at 10:46 PM John Paul Adrian Glaubitz
> >> wrote:
> >>> On 6/18/20 12:10 PM, John Paul Adrian Glaubitz wrote:
> >>> Reported up
> On Jun 19, 2020, at 10:02 AM, Geert Uytterhoeven wrote:
>
> Hi Adrian,
>
>> On Thu, Jun 18, 2020 at 10:46 PM John Paul Adrian Glaubitz
>> wrote:
>>> On 6/18/20 12:10 PM, John Paul Adrian Glaubitz wrote:
>>> Reported upstream [1].
>>
>> It's an alignment issue and can be trivially fixed w
Hi Adrian,
On Thu, Jun 18, 2020 at 10:46 PM John Paul Adrian Glaubitz
wrote:
> On 6/18/20 12:10 PM, John Paul Adrian Glaubitz wrote:
> > Reported upstream [1].
>
> It's an alignment issue and can be trivially fixed with this patch:
>
> diff --git a/op.h b/op.h
> index fc21f03cda..480c95245b 10064
Hello!
On 6/18/20 12:10 PM, John Paul Adrian Glaubitz wrote:
> Reported upstream [1].
It's an alignment issue and can be trivially fixed with this patch:
diff --git a/op.h b/op.h
index fc21f03cda..480c95245b 100644
--- a/op.h
+++ b/op.h
@@ -698,7 +698,7 @@ struct opslot {
U16 opslot
Hi!
On 6/18/20 11:58 AM, John Paul Adrian Glaubitz wrote:
>> Thanks for the heads-up. I'll look into it.
>
> Here's the backtrace:
Reported upstream [1].
Adrian
> [1] https://github.com/Perl/perl5/issues/17871
--
.''`. John Paul Adrian Glaubitz
: :' : Debian Developer - glaub...@debian.org
Hi!
On 6/17/20 10:40 PM, John Paul Adrian Glaubitz wrote:
> On 6/17/20 10:15 PM, Niko Tyni wrote:
>> perl_5.32.0~rc1-1 in experimental failed to build on m68k with a
>> SIGSEGV in miniperl. Could somebody please look at this and see if it's
>> reproducible? Guess it could be either an actual sourc
Hi!
On 6/17/20 10:15 PM, Niko Tyni wrote:
> perl_5.32.0~rc1-1 in experimental failed to build on m68k with a
> SIGSEGV in miniperl. Could somebody please look at this and see if it's
> reproducible? Guess it could be either an actual source regression or
> "just" a toolchain/platform problem.
Tha
Hi m68k porters,
perl_5.32.0~rc1-1 in experimental failed to build on m68k with a
SIGSEGV in miniperl. Could somebody please look at this and see if it's
reproducible? Guess it could be either an actual source regression or
"just" a toolchain/platform problem.
Thanks for your work,
--
Niko Tyni
18 matches
Mail list logo
