Cirrus: Automate releasing of tested binaries

It's desirable to make archives available of builds containing actual
tested content.  While not official distro-releases, these will enable
third-party testing, experimentation, and development for both branches
(e.g. "master") and pull requests (e.g. "pr3106").

* Add a Makefile targets for archiving both regular podman binaries
  and the remote-client.  Encode release metadata within these
  archives so that their exact source can be identified.

* Fix bug with cross-compiling remote clients for the Windows and Darwin
  platforms.

* Add unit-testing of cross-compiles for Windows and Darwin platforms.

* A few small CI-script typo-fixes

* Add a script which operates in two modes:

  1. Call Makefile targets which produce release archives.
     Upload the archive to Cirrus-CI's built-in caching system
     using reproducible cache keys.

  2. Utilize reproduced cache keys to attempt download of cache
     from each tasks.  When successful, parse the file's
     release metadata, using it to name the archive file.  Upload
     all recovered archives to a publicly accessible storage bucket
     for future reference.

* Update the main testing task to call the script in mode #1 for
  all primary platforms.

* Add a new `$SPECIALMODE` task to call the script in mode #1 for
  Windows and Darwin targets.

* Add a new 'release' task to the CI system, dependent upon all other
  tasks.  This new tasks executes the script in mode #2.

* Update CI documentation

Signed-off-by: Chris Evich <cevich@redhat.com>

1 files changed, 140 insertions, 0 deletions

diff --git a/contrib/cirrus/cache_release_archive.sh b/contrib/cirrus/cache_release_archive.sh
new file mode 100755
index 000000000..639bc9801
--- /dev/null
+++ b/contrib/cirrus/cache_release_archive.sh
@@ -0,0 +1,140 @@
+#!/bin/bash
+
+set -eo pipefail
+
+source $(dirname $0)/lib.sh
+
+req_env_var GOSRC
+
+RELEASE_ARCHIVE_NAMES=""
+
+handle_archive() {  # Assumed to be called with set +e
+        TASK_NUMBER=$1
+        PR_OR_BRANCH=$2
+        CACHE_URL=$3
+        ARCHIVE_NAME="$(basename $CACHE_URL)"
+        req_env_var TASK_NUMBER PR_OR_BRANCH CACHE_URL ARCHIVE_NAME
+
+        cd /tmp
+        curl -sO "$CACHE_URL" || return $(warn 0 "Couldn't download file, skipping.")
+        [[ -r "/tmp/$ARCHIVE_NAME" ]] || return $(warn 0 "Unreadable archive '/tmp/$ARCHIVE_NAME', skipping.")
+
+        ZIPCOMMENT=$(unzip -qqz "$ARCHIVE_NAME" 2>/dev/null)  # noisy bugger
+        if [[ "$?" -ne "0" ]] || [[ -z "$ZIPCOMMENT" ]]
+        then
+            return $(warn 0 "Could not unzip metadata from downloaded '/tmp/$ARCHIVE_NAME', skipping.")
+        fi
+
+        RELEASE_INFO=$(echo "$ZIPCOMMENT" | grep -m 1 'X-RELEASE-INFO:' | sed -r -e 's/X-RELEASE-INFO:\s*(.+)/\1/')
+        if [[ "$?" -ne "0" ]] || [[ -z "$RELEASE_INFO" ]]
+        then
+            return $(warn 0 "Metadata empty or invalid: '$ZIPCOMMENT', skipping.")
+        fi
+
+        # e.g. libpod v1.3.1-166-g60df124e fedora 29 amd64
+        # or   libpod v1.3.1-166-g60df124e   amd64
+        FIELDS="RELEASE_BASENAME RELEASE_VERSION RELEASE_DIST RELEASE_DIST_VER RELEASE_ARCH"
+        read $FIELDS <<< $RELEASE_INFO
+        for f in $FIELDS
+        do
+            [[ -n "${!f}" ]] || return $(warn 0 "Expecting $f to be non-empty in metadata: '$RELEASE_INFO', skipping.")
+        done
+
+        echo -n "Preparing $RELEASE_BASENAME archive: "
+        # Drop version number to enable "latest" representation
+        # (version available w/in zip-file comment)
+        RELEASE_ARCHIVE_NAME="${RELEASE_BASENAME}-${PR_OR_BRANCH}-${RELEASE_DIST}-${RELEASE_DIST_VER}-${RELEASE_ARCH}.zip"
+        # Allow uploading all gathered files in parallel, later with gsutil.
+        mv -v "$ARCHIVE_NAME" "/$RELEASE_ARCHIVE_NAME"
+        RELEASE_ARCHIVE_NAMES="$RELEASE_ARCHIVE_NAMES $RELEASE_ARCHIVE_NAME"
+}
+
+make_release() {
+    ARCHIVE_NAME="$1"
+    req_env_var ARCHIVE_NAME
+
+    # There's no actual testing of windows/darwin targets yet
+    # but we still want to cross-compile and publish binaries
+    if [[ "$SPECIALMODE" == "windows" ]] || [[ "$SPECIALMODE" == "darwin" ]]
+    then
+        RELFILE="podman-remote-${SPECIALMODE}.zip"
+    elif [[ "$SPECIALMODE" == "none" ]]
+    then
+        RELFILE="podman.zip"
+    else
+        die 55 "$(basename $0) unable to handle \$SPECIALMODE=$SPECIALMODE for $ARCHIVE_NAME"
+    fi
+    echo "Calling make $RELFILE"
+    cd $GOSRC
+    make "$RELFILE"
+    echo "Renaming archive so it can be identified/downloaded for publishing"
+    mv -v "$RELFILE" "$ARCHIVE_NAME"
+    echo "Success!"
+}
+
+[[ "$CI" == "true" ]] || \
+    die 56 "$0 requires a Cirrus-CI cross-task cache to function"
+
+cd $GOSRC
+# Same script re-used for both uploading and downloading to avoid duplication
+if [[ "$(basename $0)" == "cache_release_archive.sh" ]]
+then
+    # ref: https://cirrus-ci.org/guide/writing-tasks/#environment-variables
+    req_env_var CI_NODE_INDEX CIRRUS_BUILD_ID
+    # Use unique names for uncache_release_archives.sh to find/download them all
+    ARCHIVE_NAME="build-${CIRRUS_BUILD_ID}-task-${CI_NODE_INDEX}.zip"
+    make_release "$ARCHIVE_NAME"
+
+    # ref: https://cirrus-ci.org/guide/writing-tasks/#http-cache
+    URL="http://$CIRRUS_HTTP_CACHE_HOST/${ARCHIVE_NAME}"
+    echo "Uploading $ARCHIVE_NAME to Cirrus-CI cache at $URL"
+    curl -s -X POST --data-binary "@$ARCHIVE_NAME" "$URL"
+elif [[ "$(basename $0)" == "uncache_release_archives.sh" ]]
+then
+    req_env_var CIRRUS_BUILD_ID CI_NODE_TOTAL GCPJSON GCPNAME GCPROJECT
+    [[ "${CI_NODE_INDEX}" -eq  "$[CI_NODE_TOTAL-1]" ]] || \
+        die 8 "The release task must be executed last to guarantee archive cache is complete"
+
+    if [[ -n "$CIRRUS_PR" ]]
+    then
+        PR_OR_BRANCH="pr$CIRRUS_PR"
+        BUCKET="libpod-pr-releases"
+    elif [[ -n "$CIRRUS_BRANCH" ]]
+    then
+        PR_OR_BRANCH="$CIRRUS_BRANCH"
+        BUCKET="libpod-$CIRRUS_BRANCH-releases"
+    else
+        die 10 "Expecting either \$CIRRUS_PR or \$CIRRUS_BRANCH to be non-empty."
+    fi
+
+    echo "Blindly downloading Cirrus-CI cache files for task (some will fail)."
+    set +e  # Don't stop looping until all task's cache is attempted
+    for (( task_number = 0 ; task_number < $CI_NODE_TOTAL ; task_number++ ))
+    do
+        ARCHIVE_NAME="build-${CIRRUS_BUILD_ID}-task-${task_number}.zip"
+        URL="http://$CIRRUS_HTTP_CACHE_HOST/${ARCHIVE_NAME}"
+        echo "Attempting to download cached archive from $URL"
+        handle_archive "$task_number" "$PR_OR_BRANCH" "$URL"
+        echo "----------------------------------------"
+    done
+    set -e
+
+    [[ -n "$RELEASE_ARCHIVE_NAMES" ]] || \
+        die 67 "Error: No release archives found in CI cache, expecting at least one."
+
+    echo "Preparing to upload release archives."
+    gcloud config set project "$GCPROJECT"
+    echo "$GCPJSON" > /tmp/gcp.json
+    gcloud auth activate-service-account --key-file=/tmp/gcp.json
+    rm /tmp/gcp.json
+    # handle_archive() placed all uploadable files under /
+    gsutil -m cp /*.zip "gs://$BUCKET"  # Upload in parallel
+    echo "Successfully uploaded archives:"
+    for ARCHIVE_NAME in $RELEASE_ARCHIVE_NAMES
+    do
+        echo "    https://storage.cloud.google.com/$BUCKET/$ARCHIVE_NAME"
+    done
+    echo "These will remain available until automatic pruning by bucket policy."
+else
+    die 9 "I don't know what to do when called $0"
+fi