All of lore.kernel.org
 help / color / mirror / Atom feed
From: Weber, Matthew L                            Collins <Matthew.Weber@collins.com>
To: buildroot@busybox.net
Subject: [Buildroot] [External] Re: [PATCH v11 4/9] package/libclc: bump version to 11.1.0
Date: Wed, 19 May 2021 01:33:27 +0000	[thread overview]
Message-ID: <DM2P110MB01053EE60CA2DE6AA3196502F22B9@DM2P110MB0105.NAMP110.PROD.OUTLOOK.COM> (raw)
In-Reply-To: <dbf8d669-1d2e-4556-30f9-8a44e74942f8@mind.be>

Arnout,

> -----Original Message-----
> From: Arnout Vandecappelle <arnout@mind.be>
> Sent: Tuesday, May 18, 2021 4:04 PM
> To: Weber, Matthew L Collins <Matthew.Weber@collins.com>; buildroot
> <buildroot@buildroot.org>
> Cc: Joseph Kogut <joseph.kogut@gmail.com>; Romain Naour
> <romain.naour@gmail.com>; Samuel Martin <s.martin49@gmail.com>;
> Thomas Petazzoni <thomas.petazzoni@bootlin.com>; Valentin Korenblit
> <valentinkorenblit@gmail.com>
> Subject: [External] Re: [Buildroot] [PATCH v11 4/9] package/libclc: bump
> version to 11.1.0
> 
>  Hi Matt,
> 
>  I was going to apply this series to next...
> 
> On 11/03/2021 20:37, Matthew Weber via buildroot wrote:
> [snip]
> >>> -# C++ compiler is used to build a small tool (prepare-builtins) for the
> host.
> >>> -# It must be built with the C++ compiler from the host.
> >>> +# CMAKE_*_COMPILER_FORCED=ON skips testing the tools and
> assumes #
> >>> +llvm-config provided values
> >>>  #
> >>> -# The headers are installed in /usr/share and not /usr/include, -#
> >>> because they are needed at runtime on the target to build the OpenCL
> >>> -# kernels.
> >>> +# CMAKE_CXX_COMPILER has to be set to the host compiler to build a
> >>> +host # 'prepare_builtins' tool used during the build process
> >>>  LIBCLC_CONF_OPTS = \
> >>> -       --with-llvm-config=$(HOST_DIR)/usr/bin/llvm-config \
> >>> -       --prefix=/usr \
> >>> -       --includedir=/usr/share \
> >>> -       --pkgconfigdir=/usr/lib/pkgconfig \
> >>> -       --with-cxx-compiler=$(HOSTCXX_NOCCACHE)
> >>> +       -DCMAKE_CLC_COMPILER_FORCED=ON \
> >>> +       -DCMAKE_LLAsm_COMPILER_FORCED=ON \
> >>> +       -DCMAKE_CXX_COMPILER="$(CMAKE_HOST_CXX_COMPILER)"
> >
> > Looks like I have a couple of bugs (I didn't have opencl checked in
> > menuconfig)
> 
> ... but then I saw this. So I didn't. Instead, I've marked the series as Changes
> Requested. If you think it should be applied after all, please repost (or just
> update patchwork if unchanged).

Thx, this is definitely still broken.  I'll update the version as well when I take another look.

-Matt

> 
>  Oh, and 12.0.0 has been released now...
> 
> >
> > 1) mesa3d.mk needs MESA3D_MESON_EXTRA_BINARIES +=
> > llvm-config='$(STAGING_DIR)/bin/llvm-config' updated to point to the
> 
>  Normally, the foo-config programs are script, so it's OK to use the one in
> staging. If it's an actual binary, we have to somehow patch it so it has target
> config but compiles for host.
> 
>  I'd expect llvm and related to be cross-compile friendly though...
> 
> > HOST_DIR llvm-config because it needs to be able to execute the tool.
> > However this doesn't work since then you get the
> > "host/lib/libLLVM-11.so when searching for -lLLVM-11 " error which
> > makes sense since that's the host llvm-config we pointed at.....
> > Unsure how to solve this one and we don't notice it in x86 builds.
> > When you enable the AMD Radeon to build for aarch64 you can cause this
> > error.
> >
> > 2) mesa3d doesn't quite build (pkgconfig can't find libclc).  I didn't
> > catch that CMAKE_INSTALL_PREFIX is getting ignored (we set it in the
> > pkg-cmake.mk infra) so the libclc.pc is getting installed in
> > $HOST_DIR/usr/local/share vs $HOST_DIR/usr/share .  I haven't been
> > able to figure out why the value doesn't take hold (ie after the
> > configuration step libclc-11.1.0/CMakeCache.txt still shows /usr/local
> > unless I force set the CMAKE_INSTALL_PREFIX value in
> > libclc-11.1.0/CMakeList.txt before project().
> 
>  That *is* weird...
> 
>  Regards,
>  Arnout
> 
> >
> >
> > Matt
> > _______________________________________________
> > buildroot mailing list
> > buildroot at busybox.net
> > https://urldefense.com/v3/__http://lists.busybox.net/mailman/listinfo/
> > buildroot__;!!MvWE!SQZRb-
> v4QExXQlkcTN3XlQyQjWlvHZdLk91QeKU2DXoT1xjmvjZ
> > TjEB3aW0gMBkR77hW$
> >

  reply	other threads:[~2021-05-19  1:33 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-11 14:29 [Buildroot] [PATCH v11 1/9] package/llvm: bump to version 11.1.0 Matt Weber
2021-03-11 14:29 ` [Buildroot] [PATCH v11 2/9] package/clang: " Matt Weber
2021-03-11 14:29 ` [Buildroot] [PATCH v11 3/9] package/lld: " Matt Weber
2021-03-11 14:29 ` [Buildroot] [PATCH v11 4/9] package/libclc: bump version to 11.1.0 Matt Weber
2021-03-11 14:35   ` Matthew Weber
2021-03-11 19:37     ` Matthew Weber
2021-05-18 21:04       ` Arnout Vandecappelle
2021-05-19  1:33         ` Weber, Matthew L Collins [this message]
2022-03-22 23:25         ` James Hilliard
2021-03-11 14:29 ` [Buildroot] [PATCH v11 5/9] package/compiler-rt: new package Matt Weber
2021-03-11 14:29 ` [Buildroot] [PATCH v11 6/9] testing/tests: CLANG compiler-rt runtime test Matt Weber
2021-03-11 14:29 ` [Buildroot] [PATCH v11 7/9] package/clang: help host-clang to find our external toolchain Matt Weber
2021-03-11 14:29 ` [Buildroot] [PATCH v11 8/9] package/clang: install a toolchain-wrapper for the host clang cross-compiler Matt Weber
2021-03-11 14:29 ` [Buildroot] [PATCH v11 9/9] package/compiler-rt: runtime test use toolchain-wrapper Matt Weber

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=DM2P110MB01053EE60CA2DE6AA3196502F22B9@DM2P110MB0105.NAMP110.PROD.OUTLOOK.COM \
    --to=matthew.weber@collins.com \
    --cc=buildroot@busybox.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.