Re: [PATCH 0/5] linux-user changes to run docker

[Takashi Yamamoto]

On Tue, May 25, 2021 at 2:49 AM Alex Bennée <alex.bennee@linaro.org> wrote:
>
>
> YAMAMOTO Takashi <yamamoto@midokura.com> writes:
>
> > These patches, along with a few more hacks [1] I didn't include
> > in this patchset, allowed me to run arm64 and armv7 version of
> > dind image on amd64.
> >
> > [1] https://github.com/yamt/qemu/tree/linux-user-for-docker
>
> Might be worth posting those patches next time (even if they have a RFC
> or !MERGE in the title for now).

ok.

> I had a little noodle around with
> testing and quickly found a few holes. It would be nice if we could have
> a unit test to cover these various bits as I fear it will easily break
> again. Feel free to use the following as a basis if you want:

frankly, i feel it's enough to cover the cases which are actually used
by real apps.
is "/proc/./self/exe" etc used in the field?

>
>
>
> > You can find my test setup here:
> > https://github.com/yamt/garbage/tree/master/binfmt-aarch64-install
> >
> > YAMAMOTO Takashi (5):
> >   linux-user: handle /proc/self/exe for execve
> >   linux-uesr: make exec_path realpath
> >   linux-user: Fix the execfd case of /proc/self/exe open
> >   linux-user: dup the execfd on start up
> >   linux-user: Implement pivot_root
> >
> >  linux-user/main.c    | 14 +++++++++++++-
> >  linux-user/qemu.h    |  2 ++
> >  linux-user/syscall.c | 43 ++++++++++++++++++++++++++++++++++++++++---
> >  3 files changed, 55 insertions(+), 4 deletions(-)
>
> I also had a go at cleaning up is_proc_self and Daniel greatly
> simplified it.

thank you for the info.
unfortunately the approach seems incompatible with what i want to do
eventually. (handle non-self cases as well)

>
>
>
> --
> Alex Bennée