[U-Boot] buildman reports error of script/binutils-version.sh

[U-Boot] buildman reports error of script/binutils-version.sh

[York Sun]

Simon,

Shall we consider host error to be an error reported by buildman? I happen to
try a newer version of toolchain from Linaro. Buildman reports this error


+../scripts/binutils-version.sh: line 20: printf: 09: invalid octal number

The root cause is this version of AS reports version string differently

GNU assembler (crosstool-NG linaro-1.13.1-4.9-2014.09 - Linaro GCC 4.9-2014.09)
2.24.0.20140829 Linaro 2014.09

It has "Linaro 2014.09" at the back, so script/binutils-version.sh cannot parse
it correctly (the error is not handling zero-leading numbers, but not in the
scope of this discussion).

Another angle of this question is, why do we need this script? It doesn't sound
stable to me to parse the version string this way.

York

[U-Boot] buildman reports error of script/binutils-version.sh

[Simon Glass]

Hi,

On 21 November 2014 at 16:36, York Sun <yorksun@freescale.com> wrote:
> Simon,
>
> Shall we consider host error to be an error reported by buildman? I happen to
> try a newer version of toolchain from Linaro. Buildman reports this error
>
>
> +../scripts/binutils-version.sh: line 20: printf: 09: invalid octal number
>
> The root cause is this version of AS reports version string differently
>
> GNU assembler (crosstool-NG linaro-1.13.1-4.9-2014.09 - Linaro GCC 4.9-2014.09)
> 2.24.0.20140829 Linaro 2014.09
>
> It has "Linaro 2014.09" at the back, so script/binutils-version.sh cannot parse
> it correctly (the error is not handling zero-leading numbers, but not in the
> scope of this discussion).
>
> Another angle of this question is, why do we need this script? It doesn't sound
> stable to me to parse the version string this way.

That's another one in Masahiro's domain I think.

Regards,
Simon

[U-Boot] buildman reports error of script/binutils-version.sh

[Dirk Behme]

On 23.11.2014 23:20, Simon Glass wrote:
> Hi,
>
> On 21 November 2014 at 16:36, York Sun <yorksun@freescale.com> wrote:
>> Simon,
>>
>> Shall we consider host error to be an error reported by buildman? I happen to
>> try a newer version of toolchain from Linaro. Buildman reports this error
>>
>>
>> +../scripts/binutils-version.sh: line 20: printf: 09: invalid octal number
>>
>> The root cause is this version of AS reports version string differently
>>
>> GNU assembler (crosstool-NG linaro-1.13.1-4.9-2014.09 - Linaro GCC 4.9-2014.09)
>> 2.24.0.20140829 Linaro 2014.09
>>
>> It has "Linaro 2014.09" at the back, so script/binutils-version.sh cannot parse
>> it correctly (the error is not handling zero-leading numbers, but not in the
>> scope of this discussion).
>>
>> Another angle of this question is, why do we need this script? It doesn't sound
>> stable to me to parse the version string this way.
>
> That's another one in Masahiro's domain I think.

Are there any news/answers/fixes for this?

I'm getting above

+../scripts/binutils-version.sh: line 20: printf: 09: invalid octal number

too, with the Linaro 2014.09 tool chain.

Thanks,

Dirk

[U-Boot] buildman reports error of script/binutils-version.sh

[Masahiro Yamada]

Hi York, Dirk,


On Tue, 23 Dec 2014 15:33:05 +0100
Dirk Behme <dirk.behme@gmail.com> wrote:

> On 23.11.2014 23:20, Simon Glass wrote:
> > Hi,
> >
> > On 21 November 2014 at 16:36, York Sun <yorksun@freescale.com> wrote:
> >> Simon,
> >>
> >> Shall we consider host error to be an error reported by buildman? I happen to
> >> try a newer version of toolchain from Linaro. Buildman reports this error
> >>
> >>
> >> +../scripts/binutils-version.sh: line 20: printf: 09: invalid octal number
> >>
> >> The root cause is this version of AS reports version string differently
> >>
> >> GNU assembler (crosstool-NG linaro-1.13.1-4.9-2014.09 - Linaro GCC 4.9-2014.09)
> >> 2.24.0.20140829 Linaro 2014.09
> >>
> >> It has "Linaro 2014.09" at the back, so script/binutils-version.sh cannot parse
> >> it correctly (the error is not handling zero-leading numbers, but not in the
> >> scope of this discussion).
> >>
> >> Another angle of this question is, why do we need this script? It doesn't sound
> >> stable to me to parse the version string this way.
> >
> > That's another one in Masahiro's domain I think.
> 
> Are there any news/answers/fixes for this?
> 
> I'm getting above
> 
> +../scripts/binutils-version.sh: line 20: printf: 09: invalid octal number
> 
> too, with the Linaro 2014.09 tool chain.




Sorry for my late reply about this.


I think this patch can solve the problem
http://patchwork.ozlabs.org/patch/424032/



Best Regards
Masahiro Yamada