Re: [Buildroot] [PATCH 1/1] package/riscv-isa-sim: set --with-boost

["Yann E. MORIN" <yann.morin.1998@free.fr>]

Julien, All,

On 2022-01-23 21:03 +0100, Julien Olivain spake thusly:
> On 19/01/2022 21:23, Yann E. MORIN wrote:
[--SNIP--]
> >To be noted: riscv-isa-sim could also be configured with --without-boost
> >and the build succeeds.
> >Julien, do you know what ill-effect building without Boost would have?
> 
> The absence of Boost is removing the "Command I/O via socket" feature of
> riscv-isa-sim (-s command line option).  See:
> https://github.com/riscv-software-src/riscv-isa-sim/blob/v1.1.0/spike_main/spike.cc#L30
> It's generally not needed in simple use-cases.

Thanks for the feedback. :-)

Indeed, if Boost is not a strict requirement, it would be nice to be
able to build without it...

> My initial intent was to have host-riscv-isa-sim compiled without this Boost
> support.
> 
> The upstream riscv-isa-sim detection of boost seems a bit peculiar and does
> not
> really honor --without-boost/--without-boost-regex/--without-boost-asio.
> 
> Setting --without-boost alone on a system with boost will make the
> ./configure
> fail with:
> 
>     checking whether the Boost::ASIO library is available... yes
>     configure: error: Could not find a version of the Boost::Asio library!
> 
> Setting --without-boost --without-boost-regex --without-boost-asio will end
> up
> with a successful configuration/compilation, but with a binary linked with
> libboost_regex.so and libboost_system.so...
> 
> I believe this part could be slightly improved before setting
> --without-boost
> in the BR recipe.  I'll have a look into that.

The problem, as you said, is that if Boost is available (either already
isntalled on the build machine, or host-boost is puleld in by another
package), configuring riscv-isa-sim with --without-boost still breaks
with the message you pasted above.

But we can't guarantee that Boost (system or host-boost) is not present.

So, either we have a patch that fixes the Boost detection (and we push
that patch upstream), or we can only depend on host-boost and keep the
current status.

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 561 099 427 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot