ci: Add big endian platform

elichai commented at 1:08 pm on November 25, 2019: contributor

Discovered this as part of #17402 and a conversation with gmaxwell.

You can see here that the platform is indeed BE: https://travis-ci.org/elichai/bitcoin/jobs/616656410#L36

This closes https://github.com/bitcoin/bitcoin/issues/6466

Add ci script to install on s390x 2fa65e0de9

Add s390x tests to travis da1f153e5e

fanquake added the label Tests on Nov 25, 2019

laanwj commented at 1:10 pm on November 25, 2019: member

Concept ACK

Hope this works!

elichai commented at 1:15 pm on November 25, 2019: contributor

autotools is doing it’s job good :) https://travis-ci.org/bitcoin/bitcoin/jobs/616670043#L2809

MarcoFalke renamed this:
~~Testing bitcoin in a Big Endian platform - s390x~~
ci: AddBig Endian platform - s390x
on Nov 25, 2019

MarcoFalke renamed this:
~~ci: AddBig Endian platform - s390x~~
ci: Add big endian platform - s390x
on Nov 25, 2019

in ci/test/00_setup_env_s390x.sh:18 in da1f153e5e

13+export RUN_UNIT_TESTS=true
14+export RUN_FUNCTIONAL_TESTS=false
15+export GOAL="install"
16+export BITCOIN_CONFIG="--enable-reduce-exports --with-incompatible-bdb"
17+
18+lscpu

MarcoFalke commented at 1:33 pm on November 25, 2019:

Could add this to the script where we print free?

elichai commented at 1:40 pm on November 25, 2019:

Now that I see that it’s already printing that it’s bigendian as part of the configure I don’t mind removing this https://travis-ci.org/bitcoin/bitcoin/jobs/616670043#L2148

in ci/test/00_setup_env_s390x.sh:9 in da1f153e5e

0@@ -0,0 +1,18 @@
1+#!/usr/bin/env bash
2+#
3+# Copyright (c) 2019 The Bitcoin Core developers
4+# Distributed under the MIT software license, see the accompanying
5+# file COPYING or http://www.opensource.org/licenses/mit-license.php.
6+
7+export LC_ALL=C.UTF-8
8+
9+export HOST=s390x-unknown-linux-gnu

MarcoFalke commented at 1:36 pm on November 25, 2019:

0export HOST=s390x-unknown-linux-gnu
1# The host arch is unknown, so we run the tests through qemu.
2# If the host is arm and wants to run the tests natively, it can set QEMU_USER_CMD to the empty string.
3export QEMU_USER_CMD="${QEMU_USER_CMD:"qemu-s390x"}"

And then add in the travis.yml:

0        QEMU_USER_CMD=""  # Can run the tests natively without qemu

(just like we do for arm)

elichai commented at 1:51 pm on November 25, 2019:

So should I also add the QEMU packages that we have in the ARM script?

MarcoFalke commented at 2:00 pm on November 25, 2019:

Ah good catch, yes.

in ci/test/00_setup_env_s390x.sh:10 in da1f153e5e

 5+# file COPYING or http://www.opensource.org/licenses/mit-license.php.
 6+
 7+export LC_ALL=C.UTF-8
 8+
 9+export HOST=s390x-unknown-linux-gnu
10+export DOCKER_NAME_TAG=s390x/ubuntu:18.04

MarcoFalke commented at 1:37 pm on November 25, 2019:

Is this needed?

elichai commented at 1:42 pm on November 25, 2019:

I’ll check, in the PowerPC one Docker failed to download the right image for the platform (and instead downloaded amd64)

in ci/test/00_setup_env_s390x.sh:14 in da1f153e5e

 9+export HOST=s390x-unknown-linux-gnu
10+export DOCKER_NAME_TAG=s390x/ubuntu:18.04
11+export PACKAGES="clang llvm python3-zmq qtbase5-dev qttools5-dev-tools libssl1.0-dev libevent-dev bsdmainutils libboost-system-dev libboost-filesystem-dev libboost-chrono-dev libboost-test-dev libboost-thread-dev libdb5.3++-dev libminiupnpc-dev libzmq3-dev libqrencode-dev"
12+export NO_DEPENDS=1
13+export RUN_UNIT_TESTS=true
14+export RUN_FUNCTIONAL_TESTS=false

MarcoFalke commented at 1:37 pm on November 25, 2019:

Why?

elichai commented at 1:44 pm on November 25, 2019:

They were timing out in PowerPC, but I didn’t try them here, though because this is BE we might want to extend the timeout to make it work(because this is way more interesting then PowerPC)

MarcoFalke approved

MarcoFalke commented at 1:37 pm on November 25, 2019: member

Didn’t know that s390x is be

Thanks. Concept ACK

MarcoFalke commented at 2:01 pm on November 25, 2019: member

ACK da1f153e5e260f1744ee1bf4f24ca3a74ffea465

elichai commented at 2:11 pm on November 25, 2019: contributor

Not sure what’s up with appveyor though

MarcoFalke commented at 2:14 pm on November 25, 2019: member

Just ignore that

laanwj commented at 2:37 pm on November 25, 2019: member

yes AppVeyor issue is being fixed in #17592 and unrelated

jonatack commented at 4:43 pm on November 25, 2019: member

Concept ACK – very nice.

DrahtBot commented at 5:13 pm on November 25, 2019: member

The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.

Conflicts

Reviewers, this pull request conflicts with the following ones:

#17402 (ci: Travis support for PowerPC64 by elichai)

If you consider this pull request important, please also help to review the conflicting pull requests. Ideally, start with the one that should be merged first.

practicalswift commented at 5:16 pm on November 25, 2019: contributor

Concept ACK

in .travis.yml:106 in da1f153e5e

101@@ -102,6 +102,12 @@ jobs:
102         FILE_ENV="./ci/test/00_setup_env_arm.sh"
103         QEMU_USER_CMD=""  # Can run the tests natively without qemu
104 
105+    - stage: test
106+      name: 'S390x  [GOAL: install]  [unit tests, functional tests]'

MarcoFalke commented at 6:19 pm on November 25, 2019:

0      name: 'S390x  [GOAL: install]  [unit tests, no functional tests]'

sipa commented at 6:57 pm on November 25, 2019: member

Is there a chance of including functional tests, or is it just too slow?

MarcoFalke referenced this in commit 2c1c43754b on Nov 25, 2019

MarcoFalke merged this on Nov 25, 2019

MarcoFalke closed this on Nov 25, 2019

sidhujag referenced this in commit 8a4438f217 on Nov 25, 2019

sidhujag referenced this in commit 00ab4a1f3f on Nov 26, 2019

MarcoFalke referenced this in commit 1ae366ecb0 on Apr 9, 2020

sidhujag referenced this in commit a950439921 on Nov 10, 2020

sidhujag referenced this in commit 1b876233a4 on Nov 10, 2020

DrahtBot locked this on Feb 15, 2022

ci: Add big endian platform - s390x #17591

Conflicts