qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH] .gitlab-ci.d/windows.yml: Enable native Windows symlink


From: Thomas Huth
Subject: Re: [PATCH] .gitlab-ci.d/windows.yml: Enable native Windows symlink
Date: Wed, 27 Jul 2022 11:03:06 +0200
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101 Thunderbird/91.10.0

On 27/07/2022 10.54, Daniel P. Berrangé wrote:
On Wed, Jul 27, 2022 at 02:02:48PM +0800, Bin Meng wrote:
On Tue, Jul 26, 2022 at 9:38 AM Bin Meng <bmeng.cn@gmail.com> wrote:

On Mon, Jul 25, 2022 at 9:48 PM Alex Bennée <alex.bennee@linaro.org> wrote:


Bin Meng <bmeng.cn@gmail.com> writes:

From: Bin Meng <bin.meng@windriver.com>

The following error message was seen during the configure:

   "ln: failed to create symbolic link
   'x86_64-softmmu/qemu-system-x86_64.exe': No such file or directory"

By default the MSYS environment variable is not defined, so the runtime
behavior of winsymlinks is: if <target> does not exist, 'ln -s' fails.
At the configure phase, the qemu-system-x86_64.exe has not been built
so creation of the symbolic link fails hence the error message.

Set winsymlinks to 'native' whose behavior is most similar to the
behavior of 'ln -s' on *nix, that is:

   a) if native symlinks are enabled, and whether <target> exists
      or not, creates <destination> as a native Windows symlink;
   b) else if native symlinks are not enabled, and whether <target>
      exists or not, 'ln -s' creates as a Windows shortcut file.

Signed-off-by: Bin Meng <bin.meng@windriver.com>

I'm still seeing Windows build failures such as:

   https://gitlab.com/stsquad/qemu/-/jobs/2765579269

I've seen this one before. Looks like this one can be easily reproduced.


and

   https://gitlab.com/stsquad/qemu/-/jobs/2765579267

This one seems to be a random failure?


Saw another random failure today in the msys2-32bit build in CI.

[313/1788] Generating texture-blit-vert.h with a custom command
(wrapped by meson to capture output)
FAILED: ui/shader/texture-blit-vert.h
"C:/GitLab-Runner/builds/lbmeng/qemu/msys64/mingw32/bin/python3.exe"
"C:/GitLab-Runner/builds/lbmeng/qemu/meson/meson.py" "--internal"
"exe" "--capture" "ui/shader/texture-blit-vert.h" "--" "perl"
"C:/GitLab-Runner/builds/lbmeng/qemu/scripts/shaderinclude.pl"
"../ui/shader/texture-blit.vert"
[314/1788] Generating texture-blit-flip-vert.h with a custom command
(wrapped by meson to capture output)
ninja: build stopped: subcommand failed.
make: *** [Makefile:162: run-ninja] Error 1

IMHO we need to just drop the msys jobs from GitLab. They are too
unreliable and causing frequent failed pipelines.

IIRC they were working way more reliable 'till two or three months ago? Maybe this was introduced by the bump to the newest level of MSYS2 in commit 5c570ef2f154 ? Maybe we should upgrade or downgrade the MSYS2 version again?

 Thomas




reply via email to

[Prev in Thread] Current Thread [Next in Thread]