build-and-deploy(i686): address the dreaded .dll base address problem
#6
Conversation
|
Let's see whether this run (with the proposed fixes in effect) succeeds... |
dscho
force-pushed
the
rebaseall-in-i686-msys-builds
branch
from
8b070ef to
de7a123
Compare
Didn't. I got the wrong path for the Git for Windows SDK. Here is the next attempt. |
That failed, too, with the same Cwd.dll problem. I kicked off a re-run in the hopes that it will "magically" be resolved. If it is not, I will have to use |
|
Nope.. Apparently the Gods are not with me today. |
|
I finally think that I know what is going on. Something lets |
This is necessary to avoid the dreaded i686 problems described in https://github.com/git-for-windows/git/wiki/32-bit-issues whose symptoms can look like this: 1 [main] perl 5307 child_info_fork::abort: address space needed by 'Cwd.dll' (0x1F0000) is already occupied For an example where this symptom raises its ugly head see e.g. https://github.com/git-for-windows/git-for-windows-automation/actions/runs/3747937026/jobs/6364819440 Let's use the common strategy to run the `rebaseall` script. This is a slightly edited copy of the same step in MSYS2-packages' `build-and-deploy` workflow: https://github.com/git-for-windows/MSYS2-packages/blob/6628f78d1359/.github/workflows/build-and-deploy.yml#L60-L76 Signed-off-by: Johannes Schindelin <[email protected]>
dscho
force-pushed
the
rebaseall-in-i686-msys-builds
branch
from
7a8a04d to
fbf453b
Compare
|
We finally seem to have successfully jumped the |
dscho
force-pushed
the
rebaseall-in-i686-msys-builds
branch
from
fbf453b to
a0dda02
Compare
|
The build worked, at long last, and the i686 flavor of the So I force-pushed away the debug commits and will mark this as ready for review. |
|
Hopefully we won't need to do this anymore after the issues with Cygwin 3.4 get worked out. |
Well, Cygwin v3.4.x does not even support i686 anymore... So |
|
Right. I've somehow missed that. So the whole |
I just tried to deploy the i686 version of the MSYS variant of
curl, but ran into problems.This PR hopefully fixes those problems.