Re: libffi 3.0.11 release candidate 2 available for testing

Re: libffi 3.0.11 release candidate 2 available for testing

[Peter Kruse]

Hello,

under Solaris 10/Sparc with gcc 4.6.1 and binutils 2.21.1 (ld is
/usr/ccs/bin/ls) I get on configure:

2011-11-18 08:29:34 ./configure: line 2649: AX_ENABLE_BUILDDIR:
command not found
2011-11-18 08:29:46 ./configure: line 11842: AX_CC_MAXOPT: command not found
2011-11-18 08:29:46 ./configure: line 11843: AX_CFLAGS_WARN_ALL:
command not found
2011-11-18 08:29:49 ./configure: line 13042: GCC_AS_CFI_PSEUDO_OP:
command not found

so there is one more "command not found" since rc1.
but it does not fail.  "make check" results in:

2011-11-18 08:51:45 === libffi Summary ===
2011-11-18 08:51:45
2011-11-18 08:51:45 # of expected passes                1634
2011-11-18 08:51:45 # of unsupported tests              15

(timestamps are added by my build tool)

Regards,

  Peter

Re: libffi 3.0.11 release candidate 2 available for testing

[Anthony Green]

On 11/18/2011 2:54 AM, Peter Kruse wrote:
> Hello,
>
> under Solaris 10/Sparc with gcc 4.6.1 and binutils 2.21.1 (ld is
> /usr/ccs/bin/ls) I get on configure:
>
> 2011-11-18 08:29:34 ./configure: line 2649: AX_ENABLE_BUILDDIR:
> command not found
> 2011-11-18 08:29:46 ./configure: line 11842: AX_CC_MAXOPT: command not found
> 2011-11-18 08:29:46 ./configure: line 11843: AX_CFLAGS_WARN_ALL:
> command not found
> 2011-11-18 08:29:49 ./configure: line 13042: GCC_AS_CFI_PSEUDO_OP:
> command not found

Hi Peter,

Did you rebuild the configury or something like that?  You would get 
this if you ran "aclocal" without the "-I m4" option.

Try running "aclocal -I m4" and then autoconf again.

Thanks,

AG


>
> so there is one more "command not found" since rc1.
> but it does not fail.  "make check" results in:
>
> 2011-11-18 08:51:45 === libffi Summary ===
> 2011-11-18 08:51:45
> 2011-11-18 08:51:45 # of expected passes                1634
> 2011-11-18 08:51:45 # of unsupported tests              15
>
> (timestamps are added by my build tool)
>
> Regards,
>
>    Peter

Re: libffi 3.0.11 release candidate 2 available for testing

[Peter Kruse]

Hello Anthony,

On Fri, Nov 18, 2011 at 4:41 PM, Anthony Green <green@moxielogic.com> wrote:
> On 11/18/2011 2:54 AM, Peter Kruse wrote:
> Did you rebuild the configury or something like that?  You would get this if
> you ran "aclocal" without the "-I m4" option.

I just ran "./configure --prefix=/apps/local/gcc/4.6.1"

> Try running "aclocal -I m4" and then autoconf again.

If I do that then the result is the same.

Regards,

  Peter

libffi 3.0.11 release candidate 2 available for testing

[Anthony Green]

I've just uploaded libffi 3.0.11rc2 for testing here:
   ftp://sourceware.org/pub/libffi/libffi-3.0.11-rc2.tar.gz

I bumped the sonumber for 3.0.11 because it contains two ABI changes...

1) a debug function was removed from the non-debug builds

2) this test release includes David Gilbert's changes to support 
variadic functions.


Testing is definitely appreciated!  Please post results here:
   http://moxielogic.org/wiki/index.php?title=Libffi_3.0.11

Thanks!

Anthony Green

Re: libffi 3.0.11 release candidate 2 available for testing

[Timothy Wall]

where's the original git repo again?

On Nov 12, 2011, at 11:54 PM, Anthony Green wrote:

> I've just uploaded libffi 3.0.11rc2 for testing here:
>  ftp://sourceware.org/pub/libffi/libffi-3.0.11-rc2.tar.gz
> 
> I bumped the sonumber for 3.0.11 because it contains two ABI changes...
> 
> 1) a debug function was removed from the non-debug builds
> 
> 2) this test release includes David Gilbert's changes to support variadic functions.
> 
> 
> Testing is definitely appreciated!  Please post results here:
>  http://moxielogic.org/wiki/index.php?title=Libffi_3.0.11
> 
> Thanks!
> 
> Anthony Green
> 

Re: libffi 3.0.11 release candidate 2 available for testing

[Mark Wielaard]

On Sun, 2011-11-13 at 07:44 -0500, Timothy Wall wrote:
> where's the original git repo again?

You never know with git do you? :)
But I get mine from: git clone git://github.com/atgreen/libffi.git
Which does seem to contain the latest changes.

Cheers,

Mark

Re: libffi 3.0.11 release candidate 2 available for testing

[David Gilbert]

On 13 November 2011 04:54, Anthony Green <green@moxielogic.com> wrote:
> I've just uploaded libffi 3.0.11rc2 for testing here:
>  ftp://sourceware.org/pub/libffi/libffi-3.0.11-rc2.tar.gz
>
> I bumped the sonumber for 3.0.11 because it contains two ABI changes...
>
> 1) a debug function was removed from the non-debug builds
>
> 2) this test release includes David Gilbert's changes to support variadic
> functions.

Hi Anthony,
  Thanks for that - but it looks like a small chunk is missing from
one of the test cases:

On armhf I'm getting:

FAIL: libffi.call/cls_double_va.c -O0 -W -Wall output pattern test, is 7.0
res: 4
2.6
res: 4
? should match 7.0
?es: 4
?.0
res: 4


and looking at cls_double_va.c it's missing the chunk:

@@ -48,6 +48,10 @@ int main (void)
        printf("res: %d\n", (int) res);
        // { dg-output "\nres: 4" }

+       /* The call to cls_double_va_fn is static, so have to use a
normal prep_cif */
+       CHECK(ffi_prep_cif(&cif, FFI_DEFAULT_ABI, 2, &ffi_type_sint,
+               arg_types) == FFI_OK);
+
        CHECK(ffi_prep_closure_loc(pcl, &cif, cls_double_va_fn, NULL,
code) == FFI_OK);

        res     = ((int(*)(char*, double))(code))(format, doubleArg);

that was in my patch; the failures go away with that.

(We also get XPASS: libffi.call/cls_longdouble.c which I think is an
old one, just with it not
knowing it will work on modern ARM - that's both arm and armhf)

Dave (away for the next week)

Re: libffi 3.0.11 release candidate 2 available for testing

[Anthony Green]

On 11/18/2011 7:08 AM, David Gilbert wrote:
> On 13 November 2011 04:54, Anthony Green<green@moxielogic.com>  wrote:
>> I've just uploaded libffi 3.0.11rc2 for testing here:
>>   ftp://sourceware.org/pub/libffi/libffi-3.0.11-rc2.tar.gz
>>
>> I bumped the sonumber for 3.0.11 because it contains two ABI changes...
>>
>> 1) a debug function was removed from the non-debug builds
>>
>> 2) this test release includes David Gilbert's changes to support variadic
>> functions.
> Hi Anthony,
>    Thanks for that - but it looks like a small chunk is missing from
> one of the test cases:



Thanks for catching this David.  I didn't have the original email with 
the patch so I cut and pasted it from the web archives and it looks like 
I missed this bit.

Just FYI, there are a couple of other problems with the ARM port that 
I'd like to fix before 3.0.11.

1. It doesn't build for soft-fp targets anymore (after I added the VFP 
patches)
2. There's one macro used in the assembly code that doesn't work with 
Apple's old gas port for iOS.

Details here:  https://github.com/atgreen/libffi/issues/7

AG





>
> On armhf I'm getting:
>
> FAIL: libffi.call/cls_double_va.c -O0 -W -Wall output pattern test, is 7.0
> res: 4
> 2.6
> res: 4
> ? should match 7.0
> ?es: 4
> ?.0
> res: 4
>
>
> and looking at cls_double_va.c it's missing the chunk:
>
> @@ -48,6 +48,10 @@ int main (void)
>          printf("res: %d\n", (int) res);
>          // { dg-output "\nres: 4" }
>
> +       /* The call to cls_double_va_fn is static, so have to use a
> normal prep_cif */
> +       CHECK(ffi_prep_cif(&cif, FFI_DEFAULT_ABI, 2,&ffi_type_sint,
> +               arg_types) == FFI_OK);
> +
>          CHECK(ffi_prep_closure_loc(pcl,&cif, cls_double_va_fn, NULL,
> code) == FFI_OK);
>
>          res     = ((int(*)(char*, double))(code))(format, doubleArg);
>
> that was in my patch; the failures go away with that.
>
> (We also get XPASS: libffi.call/cls_longdouble.c which I think is an
> old one, just with it not
> knowing it will work on modern ARM - that's both arm and armhf)
>
> Dave (away for the next week)