[CI] Clone rccl and build from tip of develop (#99)

- Set cron to weekly
- Remove unused properties
- Try rccl install as sudo
- Clear existing rccl repo
- Run install with sudo and env vars
- Fix path
- Add rccl to path
- Attempt to fix build and install of rccl during compile stage.
- Remove existing clone from workspace
- Fix path when install rccl
- Fix path for install rccl-tests
- Install rccl local only
- Set RCCL_DIR
- Build rccl and rccl-tests with cmake
- Add extra env vars
- Use installer instead of cmake for rccl
- Update .jenkins/common.groovy
- Get librccl.so from rccl/build/release
- Switching job command to build rccl and rccl-tests using install.sh because those work properly together.


[ROCm/rccl-tests commit: 5c41a915c8]

projects/rccl-tests/.jenkins/common.groovy

@@ -6,17 +6,19 @@ def runCompileCommand(platform, project, jobName)
     project.paths.construct_build_prefix()
 
     String hipclangArgs = jobName.contains('hipclang') ? '--hip-clang' : ''
-    def getRCCL = auxiliary.getLibrary('rccl',platform.jenkinsLabel,'develop')
 
     def command = """#!/usr/bin/env bash
                 set -x
-                ${getRCCL}
+                cd ${project.paths.build_prefix}
+                git clone --recursive https://github.com/ROCm/rccl.git
+                cd rccl
+                ./install.sh -l
+                cd ../..
                 ${auxiliary.exitIfNotSuccess()}
+                
                 cd ${project.paths.project_build_prefix}
-                cmake \
-                    -DCMAKE_CXX_COMPILER=/opt/rocm/bin/hipcc \
-                    -S . -B build
-                make -C build -j\$(nproc)
+                export RCCL_DIR=\$(pwd)/../rccl/build/release
+                ./install.sh --rccl_home \$RCCL_DIR
                 ${auxiliary.exitIfNotSuccess()}
             """
 

projects/rccl-tests/.jenkins/precheckin.groovy

@@ -44,18 +44,16 @@ def runCI =
 ci: { 
     String urlJobName = auxiliary.getTopJobName(env.BUILD_URL)
 
-    def propertyList = ["compute-rocm-dkms-no-npi":[pipelineTriggers([cron('0 1 * * 0')])], 
-                        "compute-rocm-dkms-no-npi-hipclang":[pipelineTriggers([cron('0 1 * * 0')])],
-                        "rocm-docker":[]]
+    def propertyList = [
+        "compute-rocm-dkms-no-npi-hipclang":[pipelineTriggers([cron('0 * * * 6')])]
+    ]
     propertyList = auxiliary.appendPropertyList(propertyList)
 
-    def jobNameList = ["compute-rocm-dkms-no-npi":([ubuntu16:['rccl906']]), 
-                       "rocm-docker":([ubuntu16:['rccl906']])]
-                       
-    jobNameList['compute-rocm-dkms-no-npi-hipclang'] = [ubuntu16:['rccl906']]
+    def jobNameList = [
+        "compute-rocm-dkms-no-npi-hipclang":([ubuntu16:['rccl906']])
+    ]
     jobNameList = auxiliary.appendJobNameList(jobNameList)
     
-    
     propertyList.each 
     {
         jobName, property->
@@ -75,9 +73,9 @@ ci: {
     // For url job names that are not listed by the jobNameList i.e. compute-rocm-dkms-no-npi-1901
     if(!jobNameList.keySet().contains(urlJobName))
     {
-        properties(auxiliary.addCommonProperties([pipelineTriggers([cron('0 1 * * *')])]))
+        properties(auxiliary.addCommonProperties([pipelineTriggers([cron('0 * * * 6')])]))
         stage(urlJobName) {
             runCI([ubuntu16:['rccl906']], urlJobName)
         }
     }
-}
+}