Error building toolchain for ARM cortex A8

Error building toolchain for ARM cortex A8

[Eric B Munson]

Hi,

I am trying to build a toolchain for an ARM Cortex A8 and regardless of 
the options I select for gcc/glibc versions I always get this error:

[INFO ]  Installing C library headers & start files
[ERROR]    checking for suffix of object files... configure: error: 
cannot compute suffix of object files: cannot compile
[ERROR]
[ERROR]  >>
[ERROR]  >>  Build failed in step 'Installing C library headers & start 
files'
[ERROR]  >>        called in step '(top-level)'
[ERROR]  >>
[ERROR]  >>  Error happened in: CT_DoExecLog[scripts/functions@172]
[ERROR]  >>        called from: 
do_libc_backend_once[scripts/build/libc/glibc-eglibc.sh-common@347]
[ERROR]  >>        called from: 
do_libc_backend[scripts/build/libc/glibc-eglibc.sh-common@143]
[ERROR]  >>        called from: 
do_libc_start_files[scripts/build/libc/glibc-eglibc.sh-common@60]
[ERROR]  >>        called from: main[scripts/crosstool-NG.sh@597]
[ERROR]  >>
[ERROR]  >>  For more info on this error, look at the file: 'build.log'
[ERROR]  >>  There is a list of known issues, some with workarounds, 
in:
[ERROR]  >>      
'/usr/local/share/doc/crosstool-ng/ct-ng.hg+-9304d10c5d89/B - Known 
issues.txt'
[ERROR]
[ERROR]  (elapsed: 9:31.60)
[09:31] / make: *** [build] Error 1


The full build log is available here:

http://www.mgebm.net/~emunson/build.log

What am I doing wrong?

Thanks,
Eric

--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: Error building toolchain for ARM cortex A8

[Yann E. MORIN]

Eric, All,

On Monday 05 March 2012 22:14:02 Eric B Munson wrote:
> Hi,
> 
> I am trying to build a toolchain for an ARM Cortex A8 and regardless of 
> the options I select for gcc/glibc versions I always get this error:
> 
> [INFO ]  Installing C library headers & start files
> [ERROR]    checking for suffix of object files... configure: error: 
> cannot compute suffix of object files: cannot compile
[--SNIP--]
> The full build log is available here:
> http://www.mgebm.net/~emunson/build.log

Some time, the build.log is enough to pinpoint the issue (because usual
patterns show up, and they can be spotted).

One of the usual pattern is "See `config.log' for more details."
This points to the config.log in the current build ditrectory, which can
be inferred with the name of the step being processed, in this case, the
C library headers and start files. This points us to:
    /home/user/arm-xtool/.build/arm-unknown-linux-gnueabi/build/build-libc-startfiles/

There, a vonfig.log file shows up. Looking at this file, the last test
that was attempted and failed was:

    configure:3301: checking for suffix of object files
    configure:3327: arm-unknown-linux-gnueabi-gcc     -c  -U_FORTIFY_SOURCE
      -mlittle-endian      -mhard-float  -O2   conftest.c >&5
    conftest.c:1: sorry, unimplemented: -mfloat-abi=hard and VFP

Which looks pretty obvious: hard-float are not possible for VFP. Although
this is theorically possible, gcc-4.4 (and all other versions, AFAIK) do
not support building hard-float for VFP.

> What am I doing wrong?

Build with soft-float.

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'

--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: Error building toolchain for ARM cortex A8

[Michael Eager]

On 03/05/2012 01:47 PM, Yann E. MORIN wrote:
> Eric, All,
>
> Which looks pretty obvious: hard-float are not possible for VFP. Although
> this is theorically possible, gcc-4.4 (and all other versions, AFAIK) do
> not support building hard-float for VFP.
>
>> What am I doing wrong?
>
> Build with soft-float.

I ran into this problem as well.   There ought to be a configuration
option which selects the ARM processor version and passes the correct
options to gcc configure so that the default options do create a
broken tool chain.

-- 
Michael Eager	 eager@eagercon.com
1960 Park Blvd., Palo Alto, CA 94306  650-325-8077

--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: Error building toolchain for ARM cortex A8

[Michael Eager]

On 03/14/2012 11:10 AM, Michael Eager wrote:
> On 03/05/2012 01:47 PM, Yann E. MORIN wrote:
>> Eric, All,
>>
>> Which looks pretty obvious: hard-float are not possible for VFP. Although
>> this is theorically possible, gcc-4.4 (and all other versions, AFAIK) do
>> not support building hard-float for VFP.
>>
>>> What am I doing wrong?
>>
>> Build with soft-float.
>
> I ran into this problem as well. There ought to be a configuration
> option which selects the ARM processor version and passes the correct
> options to gcc configure so that the default options do create a
> broken tool chain.

s/do/don't/

-- 
Michael Eager	 eager@eagercon.com
1960 Park Blvd., Palo Alto, CA 94306  650-325-8077

--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: Error building toolchain for ARM cortex A8

[Michael Hope]

On 15 March 2012 07:10, Michael Eager <eager@eagerm.com> wrote:
> On 03/05/2012 01:47 PM, Yann E. MORIN wrote:
>>
>> Eric, All,
>>
>> Which looks pretty obvious: hard-float are not possible for VFP. Although
>> this is theorically possible, gcc-4.4 (and all other versions, AFAIK) do
>> not support building hard-float for VFP.
>>
>>> What am I doing wrong?
>>
>>
>> Build with soft-float.
>
>
> I ran into this problem as well.   There ought to be a configuration
> option which selects the ARM processor version and passes the correct
> options to gcc configure so that the default options do create a
> broken tool chain.

Sorry, I'm late to this thread.  GCC 4.5 and later support the hard
float ABI on the Cortex-As with VFPv3.  I think 4.4 does as well.
What was the problem?

-- Michael

--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: Error building toolchain for ARM cortex A8

[Bryan Hundven]

On Wed, Mar 14, 2012 at 1:36 PM, Michael Hope <michael.hope@linaro.org> wrote:
> On 15 March 2012 07:10, Michael Eager <eager@eagerm.com> wrote:
>> On 03/05/2012 01:47 PM, Yann E. MORIN wrote:
>>>
>>> Eric, All,
>>>
>>> Which looks pretty obvious: hard-float are not possible for VFP. Although
>>> this is theorically possible, gcc-4.4 (and all other versions, AFAIK) do
>>> not support building hard-float for VFP.
>>>
>>>> What am I doing wrong?
>>>
>>>
>>> Build with soft-float.
>>
>>
>> I ran into this problem as well.   There ought to be a configuration
>> option which selects the ARM processor version and passes the correct
>> options to gcc configure so that the default options do create a
>> broken tool chain.
>
> Sorry, I'm late to this thread.  GCC 4.5 and later support the hard
> float ABI on the Cortex-As with VFPv3.  I think 4.4 does as well.
> What was the problem?

He's building 4.4.6.

> -- Michael
>
> --
> For unsubscribe information see http://sourceware.org/lists.html#faq
>

--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: Error building toolchain for ARM cortex A8

[Michael Eager]

On 03/14/2012 01:40 PM, Bryan Hundven wrote:
> On Wed, Mar 14, 2012 at 1:36 PM, Michael Hope<michael.hope@linaro.org>  wrote:
>> On 15 March 2012 07:10, Michael Eager<eager@eagerm.com>  wrote:
>>> On 03/05/2012 01:47 PM, Yann E. MORIN wrote:
>>>>
>>>> Eric, All,
>>>>
>>>> Which looks pretty obvious: hard-float are not possible for VFP. Although
>>>> this is theorically possible, gcc-4.4 (and all other versions, AFAIK) do
>>>> not support building hard-float for VFP.
>>>>
>>>>> What am I doing wrong?
>>>>
>>>>
>>>> Build with soft-float.
>>>
>>>
>>> I ran into this problem as well.   There ought to be a configuration
>>> option which selects the ARM processor version and passes the correct
>>> options to gcc configure so that the default options do create a
>>> broken tool chain.
>>
>> Sorry, I'm late to this thread.  GCC 4.5 and later support the hard
>> float ABI on the Cortex-As with VFPv3.  I think 4.4 does as well.
>> What was the problem?
>
> He's building 4.4.6.

I saw the scripts/addToolVersion.sh script and added gcc-4.6.3.
But when I run "ct-ng menuconfig" this is not an option for GCC.
How do I make this show up in menuconfig?

-- 
Michael Eager	 eager@eagercon.com
1960 Park Blvd., Palo Alto, CA 94306  650-325-8077

--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: Error building toolchain for ARM cortex A8

[Bryan Hundven]

On Wed, Mar 14, 2012 at 3:50 PM, Michael Eager <eager@eagerm.com> wrote:
> On 03/14/2012 01:40 PM, Bryan Hundven wrote:
>>
>> On Wed, Mar 14, 2012 at 1:36 PM, Michael Hope<michael.hope@linaro.org>
>>  wrote:
>>>
>>> On 15 March 2012 07:10, Michael Eager<eager@eagerm.com>  wrote:
>>>>
>>>> On 03/05/2012 01:47 PM, Yann E. MORIN wrote:
>>>>>
>>>>>
>>>>> Eric, All,
>>>>>
>>>>> Which looks pretty obvious: hard-float are not possible for VFP.
>>>>> Although
>>>>> this is theorically possible, gcc-4.4 (and all other versions, AFAIK)
>>>>> do
>>>>> not support building hard-float for VFP.
>>>>>
>>>>>> What am I doing wrong?
>>>>>
>>>>>
>>>>>
>>>>> Build with soft-float.
>>>>
>>>>
>>>>
>>>> I ran into this problem as well.   There ought to be a configuration
>>>> option which selects the ARM processor version and passes the correct
>>>> options to gcc configure so that the default options do create a
>>>> broken tool chain.
>>>
>>>
>>> Sorry, I'm late to this thread.  GCC 4.5 and later support the hard
>>> float ABI on the Cortex-As with VFPv3.  I think 4.4 does as well.
>>> What was the problem?
>>
>>
>> He's building 4.4.6.
>
>
> I saw the scripts/addToolVersion.sh script and added gcc-4.6.3.
> But when I run "ct-ng menuconfig" this is not an option for GCC.
> How do I make this show up in menuconfig?

You need CT_EXPERIMENTAL=y

>
> --
> Michael Eager    eager@eagercon.com
> 1960 Park Blvd., Palo Alto, CA 94306  650-325-8077

--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: Error building toolchain for ARM cortex A8

[Michael Eager]

On 03/14/2012 04:06 PM, Bryan Hundven wrote:
> On Wed, Mar 14, 2012 at 3:50 PM, Michael Eager<eager@eagerm.com>  wrote:
>> On 03/14/2012 01:40 PM, Bryan Hundven wrote:
>>>
>>> On Wed, Mar 14, 2012 at 1:36 PM, Michael Hope<michael.hope@linaro.org>
>>>   wrote:
>>>>
>>>> On 15 March 2012 07:10, Michael Eager<eager@eagerm.com>    wrote:
>>>>>
>>>>> On 03/05/2012 01:47 PM, Yann E. MORIN wrote:
>>>>>>
>>>>>>
>>>>>> Eric, All,
>>>>>>
>>>>>> Which looks pretty obvious: hard-float are not possible for VFP.
>>>>>> Although
>>>>>> this is theorically possible, gcc-4.4 (and all other versions, AFAIK)
>>>>>> do
>>>>>> not support building hard-float for VFP.
>>>>>>
>>>>>>> What am I doing wrong?
>>>>>>
>>>>>>
>>>>>>
>>>>>> Build with soft-float.
>>>>>
>>>>>
>>>>>
>>>>> I ran into this problem as well.   There ought to be a configuration
>>>>> option which selects the ARM processor version and passes the correct
>>>>> options to gcc configure so that the default options do create a
>>>>> broken tool chain.
>>>>
>>>>
>>>> Sorry, I'm late to this thread.  GCC 4.5 and later support the hard
>>>> float ABI on the Cortex-As with VFPv3.  I think 4.4 does as well.
>>>> What was the problem?
>>>
>>>
>>> He's building 4.4.6.
>>
>>
>> I saw the scripts/addToolVersion.sh script and added gcc-4.6.3.
>> But when I run "ct-ng menuconfig" this is not an option for GCC.
>> How do I make this show up in menuconfig?
>
> You need CT_EXPERIMENTAL=y

I ran "CT_EXPERIMENTAL=y ct-ng menuconfig" but I don't see gcc-4.6.3.

-- 
Michael Eager	 eager@eagercon.com
1960 Park Blvd., Palo Alto, CA 94306  650-325-8077

--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: Error building toolchain for ARM cortex A8

[Richard Earnshaw]

On 14/03/12 20:36, Michael Hope wrote:
> On 15 March 2012 07:10, Michael Eager <eager@eagerm.com> wrote:
>> On 03/05/2012 01:47 PM, Yann E. MORIN wrote:
>>>
>>> Eric, All,
>>>
>>> Which looks pretty obvious: hard-float are not possible for VFP. Although
>>> this is theorically possible, gcc-4.4 (and all other versions, AFAIK) do
>>> not support building hard-float for VFP.
>>>
>>>> What am I doing wrong?
>>>
>>>
>>> Build with soft-float.
>>
>>
>> I ran into this problem as well.   There ought to be a configuration
>> option which selects the ARM processor version and passes the correct
>> options to gcc configure so that the default options do create a
>> broken tool chain.
> 
> Sorry, I'm late to this thread.  GCC 4.5 and later support the hard
> float ABI on the Cortex-As with VFPv3.  I think 4.4 does as well.
> What was the problem?
> 

Hard float ABI was new in gcc-4.5.  There's a back-port to 4.4 in
svn://gcc.gnu.org/svn/gcc/branches/ARM/hard_vfp_4_4_branch but it's
probably not very up-to-date relative to the official 4.4 branch.

R.


--
For unsubscribe information see http://sourceware.org/lists.html#faq