[OE-core] [RFT][PATCH] gcc-9.0: Add recipes for upcoming gcc 9.0 release in mid-2019

richard.purdie at linuxfoundation.org richard.purdie at linuxfoundation.org
Thu May 2 09:38:16 UTC 2019


On Wed, 2019-05-01 at 20:01 -0700, Khem Raj wrote:
> On Wed, May 1, 2019 at 4:34 AM Richard Purdie
> <richard.purdie at linuxfoundation.org> wrote:
> > On Tue, 2019-04-30 at 12:24 -0700, Khem Raj wrote:
> > > Its already in RC phase now.
> > > 
> > > Signed-off-by: Khem Raj <raj.khem at gmail.com>
> > 
> > I ran this on the autobuilder:
> > 
> > https://autobuilder.yoctoproject.org/typhoon/#/builders/85/builds/429
> > http://errors.yoctoproject.org/Errors/Details/238976/
> > 
> > Basic summary:
> > 
> > a) gcc-source needs a maintainers.inc entry (selftest fail)
> 
> I will fix that in next rev.
> 
> > b) fortran selftest is broken
> 
> some more details on this would be helpful.

"oe-selftest -r buildoptions.ToolchainOptions.test_toolchain_fortran"
should reproduce.

| In file included from ../../../../../../../../work-shared/gcc-9.0.1-r0/official-gcc-7833ad1/libgfortran/runtime/backtrace.c:27:
| ../../../../../../../../work-shared/gcc-9.0.1-r0/official-gcc-7833ad1/libgfortran/../libgcc/gthr.h:148:10: fatal error: gthr-default.h: No such file or directory
|  #include "gthr-default.h"
|           ^~~~~~~~~~~~~~~~
| compilation terminated.
| Makefile:2772: recipe for target 'backtrace.lo' failed

from

https://autobuilder.yoctoproject.org/typhoon/api/v2/logs/468206/raw

> 
> > c) mdadm fails to compile
> 
> Yes, this appeared in my builds too but got ignored since recent
> builds are using clang
> but I will look into fixing it.
> 
> > d) runtime test gpp2 fails
> > e) systemd fails to compile
> > f) on target runtime g++ compile tests fail
> 
> The headers include problem is interesting I will see if I can
> reproduce it.

"bitbake core-image-sato-sdk -c testimage"

should reproduce.

> > Build will look a lot happier with the on target C++ headers fixed! Not
> > too bad for a first pass :)
>
> This is because we can build master branch easily for toolchain
> recipes. so a lot of pre-emptive work has happened already

I can imagine there was quite some work in that, nicely done! :)

Cheers,

Richard





More information about the Openembedded-core mailing list