[PATCH 3/6] shgen: Import from RTEMS
Gedare Bloom
gedare at rtems.org
Mon Jun 11 14:21:43 UTC 2018
On Mon, Jun 11, 2018 at 5:35 AM, Chris Johns <chrisj at rtems.org> wrote:
> On 11/6/18 5:49 pm, Sebastian Huber wrote:
>> On 11/06/18 08:22, Joel Sherrill wrote:
>>>
>>>
>>> >
>>> > >
>>> > >>> Does this tool need to move over?
>>> > >> I would like to move all host tools out of the RTEMS
>>> repository
>>> > to simplify the
>>> > >> build. If we get rid of the host tools, then we have only to
>>> > deal with the
>>> > >> cross-build tools.
>>> > > This is a good thing to do, thank you.
>>> > >
>>> > >>> Is the sh arch active?
>>> > >> I don't know.
>>> > > I suggest we remove this code until someone needs it and can
>>> > resolve this issue.
>>> > > The code is in the 4.11 releases so is easy to find.
>>> >
>>> > The shgen tool is used to build the gensh1 and gensh2 BSPs to
>>> > generate
>>> > the scitab.c source file.
>>>
>>>
>>> Sorry..I missed this.
>>>
>>> >
>>> > One option to get rid of the shgen tool would be to remove the
>>> > CPU_CLOCK_RATE_HZ BSP option.
>>>
>>>
>>> Ok. Or retroactively deprecate in 4.11
>>
>> The tool generates a scitab.c with this header:
>>
>> /*
>> * DO NOT EDIT - this file is automatically generated by shgen
>> 5.07adc7f7e37a_modified
>> * Copyright (c) 1998,1999,2000 Ralf Corsepius (corsepiu at faw.uni-ulm.de)
>> */
>>
>> /* This file is not copyrighted */
>>
>> Files with such a header are not acceptable for an import into the RTEMS sources
>> repository from my point of view.
>
> I agree.
>
>> We have three options now:
>>
>> 1. Keep the tools in the RTEMS source repository.
>
> I prefer they are moved out.
>
>> 2. Add GPL software to rtems-tools.
>
> I prefer this not happen. I have worked hard over the life of repo to be careful
> not to have GPL in the repo.
>
Option 2.b: Create contrib-tools repo with third party licensed softwares.
>>
>> 3. Remove the gensh1 and gensh2 BSPs.
>>
>
> What is in the generated file?
>
> We use weak symbols in the Zynq BSP to allow us to override defaults. Could this
> approach be used?
>
> Chris
> _______________________________________________
> devel mailing list
> devel at rtems.org
> http://lists.rtems.org/mailman/listinfo/devel
More information about the devel
mailing list