Jenkinsfile 31.2 KB
Newer Older
JD's avatar
JD committed
1
2
3
4
5
6
7
8
9
10
11
12
13
def rocmnode(name) {
    return 'rocmtest && miopen && ' + name
}

def show_node_info() {
    sh """
        echo "NODE_NAME = \$NODE_NAME"
        lsb_release -sd
        uname -r
        ls /opt/ -la
    """
}

14
def runShell(String command){
15
    def responseCode = sh returnStatus: true, script: "${command} > tmp.txt"
16
    def output = readFile(file: "tmp.txt")
17
    echo "tmp.txt contents: $output"
18
19
20
    return (output != "")
}

21
def getDockerImageName(){
22
    def img = "${env.CK_IMAGE_URL}:ck_ub20.04_rocm5.2.3_${params.COMPILER_VERSION}"
23
24
25
    return img
}

26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
def build_compiler(){
    def compiler
    if (params.BUILD_COMPILER == "hipcc"){
        compiler = '/opt/rocm/bin/hipcc'
    }
    else{
        if (params.COMPILER_VERSION == "release"){
            compiler = "/opt/rocm/llvm/bin/clang++"
        }
        else{
            compiler = "/llvm-project/build/bin/clang++"
        }        
    }
    return compiler
}

42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
def getDockerImage(Map conf=[:]){
    env.DOCKER_BUILDKIT=1
    def prefixpath = conf.get("prefixpath", "/opt/rocm") // prefix:/opt/rocm
    def no_cache = conf.get("no_cache", false)
    def dockerArgs = "--build-arg BUILDKIT_INLINE_CACHE=1 --build-arg PREFIX=${prefixpath} --build-arg compiler_version='${params.COMPILER_VERSION}' "
    if(env.CCACHE_HOST)
    {
        def check_host = sh(script:"""(printf "PING\r\n";) | nc -N ${env.CCACHE_HOST} 6379 """, returnStdout: true).trim()
        if(check_host == "+PONG")
        {
            echo "FOUND CCACHE SERVER: ${CCACHE_HOST}"
        }
        else 
        {
            echo "CCACHE SERVER: ${CCACHE_HOST} NOT FOUND, got ${check_host} response"
        }
        dockerArgs = dockerArgs + " --build-arg CCACHE_SECONDARY_STORAGE='redis://${env.CCACHE_HOST}' --build-arg COMPILER_LAUNCHER='ccache' "
        env.CCACHE_DIR = """/tmp/ccache_store"""
        env.CCACHE_SECONDARY_STORAGE="""redis://${env.CCACHE_HOST}"""
    }
    if(no_cache)
    {
        dockerArgs = dockerArgs + " --no-cache "
    }
    echo "Docker Args: ${dockerArgs}"
    def image = getDockerImageName()
    //Check if image exists 
    def retimage
    try 
    {
        echo "Pulling down image: ${image}"
        retimage = docker.image("${image}")
        retimage.pull()
    }
    catch(Exception ex)
    {
        error "Unable to locate image: ${image}"
    }
    return [retimage, image]
}

def buildDocker(install_prefix){
    show_node_info()
    env.DOCKER_BUILDKIT=1
    checkout scm
    def image_name = getDockerImageName()
    echo "Building Docker for ${image_name}"
    def dockerArgs = "--build-arg BUILDKIT_INLINE_CACHE=1 --build-arg PREFIX=${install_prefix} --build-arg compiler_version='${params.COMPILER_VERSION}' "
    if(env.CCACHE_HOST)
    {
        def check_host = sh(script:"""(printf "PING\\r\\n";) | nc  -N ${env.CCACHE_HOST} 6379 """, returnStdout: true).trim()
        if(check_host == "+PONG")
        {
            echo "FOUND CCACHE SERVER: ${CCACHE_HOST}"
        }
        else 
        {
            echo "CCACHE SERVER: ${CCACHE_HOST} NOT FOUND, got ${check_host} response"
        }
        dockerArgs = dockerArgs + " --build-arg CCACHE_SECONDARY_STORAGE='redis://${env.CCACHE_HOST}' --build-arg COMPILER_LAUNCHER='ccache' "
        env.CCACHE_DIR = """/tmp/ccache_store"""
        env.CCACHE_SECONDARY_STORAGE="""redis://${env.CCACHE_HOST}"""
    }

    echo "Build Args: ${dockerArgs}"
    try{
        echo "Checking for image: ${image_name}"
        sh "docker manifest inspect --insecure ${image_name}"
        echo "Image: ${image_name} found!! Skipping building image"
    }
    catch(Exception ex){
        echo "Unable to locate image: ${image_name}. Building image now"
        retimage = docker.build("${image_name}", dockerArgs + ' .')
        retimage.push()
    }
}

JD's avatar
JD committed
119
120
def cmake_build(Map conf=[:]){

121
    def compiler = build_compiler()
JD's avatar
JD committed
122
123
    def config_targets = conf.get("config_targets","check")
    def debug_flags = "-g -fno-omit-frame-pointer -fsanitize=undefined -fno-sanitize-recover=undefined " + conf.get("extradebugflags", "")
124
    def build_envs = "CTEST_PARALLEL_LEVEL=4 " + conf.get("build_env","")
JD's avatar
JD committed
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
    def prefixpath = conf.get("prefixpath","/opt/rocm")
    def setup_args = conf.get("setup_args","")

    if (prefixpath != "/usr/local"){
        setup_args = setup_args + " -DCMAKE_PREFIX_PATH=${prefixpath} "
    }

    def build_type_debug = (conf.get("build_type",'release') == 'debug')

    //cmake_env can overwrite default CXX variables.
    def cmake_envs = "CXX=${compiler} CXXFLAGS='-Werror' " + conf.get("cmake_ex_env","")

    def package_build = (conf.get("package_build","") == "true")

    if (package_build == true) {
        config_targets = "package"
    }

    if(conf.get("build_install","") == "true")
    {
        config_targets = 'install ' + config_targets
        setup_args = ' -DBUILD_DEV=Off -DCMAKE_INSTALL_PREFIX=../install' + setup_args
    } else{
        setup_args = ' -DBUILD_DEV=On' + setup_args
    }

    if(build_type_debug){
        setup_args = " -DCMAKE_BUILD_TYPE=debug -DCMAKE_CXX_FLAGS_DEBUG='${debug_flags}'" + setup_args
    }else{
        setup_args = " -DCMAKE_BUILD_TYPE=release" + setup_args
    }
156
157
158
159
    if(env.CCACHE_HOST)
    {
        setup_args = " -DCMAKE_CXX_COMPILER_LAUNCHER='ccache' -DCMAKE_C_COMPILER_LAUNCHER='ccache' " + setup_args
    }
JD's avatar
JD committed
160
161
162
163
164
165
166
167
168
169
170

    def pre_setup_cmd = """
            echo \$HSA_ENABLE_SDMA
            ulimit -c unlimited
            rm -rf build
            mkdir build
            rm -rf install
            mkdir install
            cd build
        """
    def setup_cmd = conf.get("setup_cmd", "${cmake_envs} cmake ${setup_args}   .. ")
Chao Liu's avatar
Chao Liu committed
171
    // reduce parallelism when compiling, clang uses too much memory
172
    def build_cmd = conf.get("build_cmd", "${build_envs} dumb-init make  -j\$(( \$(nproc) / 2 )) ${config_targets}")
JD's avatar
JD committed
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
    def execute_cmd = conf.get("execute_cmd", "")

    def cmd = conf.get("cmd", """
            ${pre_setup_cmd}
            ${setup_cmd}
            ${build_cmd}
            ${execute_cmd}
        """)

    echo cmd
    sh cmd

    // Only archive from master or develop
    if (package_build == true && (env.BRANCH_NAME == "develop" || env.BRANCH_NAME == "master")) {
        archiveArtifacts artifacts: "build/*.deb", allowEmptyArchive: true, fingerprint: true
    }
}

def buildHipClangJob(Map conf=[:]){
        show_node_info()

        env.HSA_ENABLE_SDMA=0
        checkout scm

197
        def image = getDockerImageName() 
JD's avatar
JD committed
198
199
200
        def prefixpath = conf.get("prefixpath", "/opt/rocm")

        // Jenkins is complaining about the render group 
201
        def dockerOpts="--device=/dev/kfd --device=/dev/dri --group-add video --group-add render --cap-add=SYS_PTRACE --security-opt seccomp=unconfined"
JD's avatar
JD committed
202
        if (conf.get("enforce_xnack_on", false)) {
203
            dockerOpts = dockerOpts + " --env HSA_XNACK=1 "
JD's avatar
JD committed
204
        }
205
        def dockerArgs = "--build-arg PREFIX=${prefixpath} --build-arg compiler_version='${params.COMPILER_VERSION}' "
206
        if (params.COMPILER_VERSION != "release"){
207
208
            dockerOpts = dockerOpts + " --env HIP_CLANG_PATH='/llvm-project/build/bin' "
        }
JD's avatar
JD committed
209
210
211
212

        def variant = env.STAGE_NAME

        def retimage
213
214

        gitStatusWrapper(credentialsId: "${status_wrapper_creds}", gitHubContext: "Jenkins - ${variant}", account: 'ROCmSoftwarePlatform', repo: 'composable_kernel') {
215
            try {
216
217
                //retimage = docker.build("${image}", dockerArgs + '.')
                (retimage, image) = getDockerImage(conf)
218
219
                withDockerContainer(image: image, args: dockerOpts) {
                    timeout(time: 5, unit: 'MINUTES'){
220
221
                        sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo | tee clinfo.log'
                        if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
222
                            throw new Exception ("GPU not found")
223
224
225
226
                        }
                        else{
                            echo "GPU is OK"
                        }
JD's avatar
JD committed
227
228
229
                    }
                }
            }
230
231
232
233
234
235
236
237
            catch (org.jenkinsci.plugins.workflow.steps.FlowInterruptedException e){
                echo "The job was cancelled or aborted"
                throw e
            }
            catch(Exception ex) {
                retimage = docker.build("${image}", dockerArgs + " --no-cache .")
                withDockerContainer(image: image, args: dockerOpts) {
                    timeout(time: 5, unit: 'MINUTES'){
238
239
                        sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo |tee clinfo.log'
                        if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
240
                            throw new Exception ("GPU not found")
241
242
243
244
                        }
                        else{
                            echo "GPU is OK"
                        }
245
                    }
246
247
                }
            }
JD's avatar
JD committed
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278

            withDockerContainer(image: image, args: dockerOpts + ' -v=/var/jenkins/:/var/jenkins') {
                timeout(time: 5, unit: 'HOURS')
                {
                    cmake_build(conf)
                }
            }
        }
        return retimage
}

def reboot(){
    build job: 'reboot-slaves', propagate: false , parameters: [string(name: 'server', value: "${env.NODE_NAME}"),]
}

def buildHipClangJobAndReboot(Map conf=[:]){
    try{
        buildHipClangJob(conf)
    }
    catch(e){
        echo "throwing error exception for the stage"
        echo 'Exception occurred: ' + e.toString()
        throw e
    }
    finally{
        if (!conf.get("no_reboot", false)) {
            reboot()
        }
    }
}

279
280
281
282
283
284
def runCKProfiler(Map conf=[:]){
        show_node_info()

        env.HSA_ENABLE_SDMA=0
        checkout scm

285
        def image = getDockerImageName()
286
287
288
        def prefixpath = conf.get("prefixpath", "/opt/rocm")

        // Jenkins is complaining about the render group 
289
        def dockerOpts="--device=/dev/kfd --device=/dev/dri --group-add video --group-add render --cap-add=SYS_PTRACE --security-opt seccomp=unconfined"
290
        if (conf.get("enforce_xnack_on", false)) {
291
            dockerOpts = dockerOpts + " --env HSA_XNACK=1 "
292
        }
293
        def dockerArgs = "--build-arg PREFIX=${prefixpath} --build-arg compiler_version='${params.COMPILER_VERSION}' "
294
        if (params.COMPILER_VERSION != "release"){
295
296
            dockerOpts = dockerOpts + " --env HIP_CLANG_PATH='/llvm-project/build/bin' "
        }
297
298
299

        def variant = env.STAGE_NAME
        def retimage
300
301

        gitStatusWrapper(credentialsId: "${status_wrapper_creds}", gitHubContext: "Jenkins - ${variant}", account: 'ROCmSoftwarePlatform', repo: 'composable_kernel') {
302
            try {
303
                (retimage, image) = getDockerImage(conf)
304
305
                withDockerContainer(image: image, args: dockerOpts) {
                    timeout(time: 5, unit: 'MINUTES'){
306
307
                        sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo | tee clinfo.log'
                        if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
308
                            throw new Exception ("GPU not found")
309
310
311
312
                        }
                        else{
                            echo "GPU is OK"
                        }
313
314
315
                    }
                }
            }
316
317
318
319
320
321
322
323
            catch (org.jenkinsci.plugins.workflow.steps.FlowInterruptedException e){
                echo "The job was cancelled or aborted"
                throw e
            }
            catch(Exception ex) {
                retimage = docker.build("${image}", dockerArgs + " --no-cache .")
                withDockerContainer(image: image, args: dockerOpts) {
                    timeout(time: 5, unit: 'MINUTES'){
324
325
                        sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo | tee clinfo.log'
                        if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
326
                            throw new Exception ("GPU not found")
327
328
329
330
                        }
                        else{
                            echo "GPU is OK"
                        }
331
                    }
332
333
                }
            }
334
335

            withDockerContainer(image: image, args: dockerOpts + ' -v=/var/jenkins/:/var/jenkins') {
336
                timeout(time: 24, unit: 'HOURS')
337
                {
338
339
340
341
342
343
344
345
346
347
348
                    //cmake_build(conf)
                    //instead of building, just unstash the ckProfiler and install it
                    sh """
                        rm -rf build
                        mkdir build
                    """
                    dir("build"){
                        unstash 'ckProfiler.tar.gz'
                        sh 'tar -xvf ckProfiler.tar.gz'
                    }

349
					dir("script"){
350
                        if (params.RUN_FULL_QA){
351
352
353
354
355
356
357
358
359
360
361
362
363
                            sh "./run_full_performance_tests.sh 1 QA_${params.COMPILER_VERSION} ${env.BRANCH_NAME} ${NODE_NAME}"
                            archiveArtifacts "perf_gemm.log"
                            archiveArtifacts "perf_resnet50_N256.log"
                            archiveArtifacts "perf_resnet50_N4.log"
                            archiveArtifacts "perf_batched_gemm.log"
                            archiveArtifacts "perf_grouped_gemm.log"
                            archiveArtifacts "perf_conv_fwd.log"
                            archiveArtifacts "perf_conv_bwd_data.log"
                            archiveArtifacts "perf_gemm_bilinear.log"
                            archiveArtifacts "perf_reduction.log"
                            archiveArtifacts "perf_splitK_gemm_verify.log"
                            archiveArtifacts "perf_splitK_gemm.log"
                            archiveArtifacts "perf_onnx_gemm.log"
364
                           // stash perf files to master
365
366
367
368
369
370
371
372
373
374
375
                            stash name: "perf_gemm.log"
                            stash name: "perf_resnet50_N256.log"
                            stash name: "perf_resnet50_N4.log"
                            stash name: "perf_batched_gemm.log"
                            stash name: "perf_grouped_gemm.log"
                            stash name: "perf_conv_fwd.log"
                            stash name: "perf_conv_bwd_data.log"
                            stash name: "perf_gemm_bilinear.log"
                            stash name: "perf_reduction.log"
                            stash name: "perf_splitK_gemm.log"
                            stash name: "perf_onnx_gemm.log"
376
                            //we will process results on the master node
377
378
                        }
                        else{
379
380
381
382
                            sh "./run_performance_tests.sh 0 CI_${params.COMPILER_VERSION} ${env.BRANCH_NAME} ${NODE_NAME}"
                            archiveArtifacts "perf_gemm.log"
                            archiveArtifacts "perf_resnet50_N256.log"
                            archiveArtifacts "perf_resnet50_N4.log"
383
                            // stash perf files to master
384
385
386
                            stash name: "perf_gemm.log"
                            stash name: "perf_resnet50_N256.log"
                            stash name: "perf_resnet50_N4.log"
387
                            //we will process the results on the master node
388
                        }
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
					}
                }
            }
        }
        return retimage
}

def runPerfTest(Map conf=[:]){
    try{
        runCKProfiler(conf)
    }
    catch(e){
        echo "throwing error exception in performance tests"
        echo 'Exception occurred: ' + e.toString()
        throw e
    }
    finally{
        if (!conf.get("no_reboot", false)) {
            reboot()
        }
    }
}

412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
def Build_CK(Map conf=[:]){
        show_node_info()

        env.HSA_ENABLE_SDMA=0
        checkout scm

        def image = getDockerImageName() 
        def prefixpath = conf.get("prefixpath", "/opt/rocm")

        // Jenkins is complaining about the render group 
        def dockerOpts="--device=/dev/kfd --device=/dev/dri --group-add video --group-add render --cap-add=SYS_PTRACE --security-opt seccomp=unconfined"
        if (conf.get("enforce_xnack_on", false)) {
            dockerOpts = dockerOpts + " --env HSA_XNACK=1 "
        }
        def dockerArgs = "--build-arg PREFIX=${prefixpath} --build-arg compiler_version='${params.COMPILER_VERSION}' "
        if (params.COMPILER_VERSION != "release"){
            dockerOpts = dockerOpts + " --env HIP_CLANG_PATH='/llvm-project/build/bin' "
        }

        def variant = env.STAGE_NAME
        def retimage

        gitStatusWrapper(credentialsId: "${status_wrapper_creds}", gitHubContext: "Jenkins - ${variant}", account: 'ROCmSoftwarePlatform', repo: 'composable_kernel') {
            try {
                (retimage, image) = getDockerImage(conf)
                withDockerContainer(image: image, args: dockerOpts) {
                    timeout(time: 5, unit: 'MINUTES'){
                        sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo | tee clinfo.log'
                        if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
                            throw new Exception ("GPU not found")
                        }
                        else{
                            echo "GPU is OK"
                        }
                    }
                }
            }
            catch (org.jenkinsci.plugins.workflow.steps.FlowInterruptedException e){
                echo "The job was cancelled or aborted"
                throw e
            }
            catch(Exception ex) {
                retimage = docker.build("${image}", dockerArgs + " --no-cache .")
                withDockerContainer(image: image, args: dockerOpts) {
                    timeout(time: 5, unit: 'MINUTES'){
                        sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo |tee clinfo.log'
                        if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
                            throw new Exception ("GPU not found")
                        }
                        else{
                            echo "GPU is OK"
                        }
                    }
                }
            }
            withDockerContainer(image: image, args: dockerOpts + ' -v=/var/jenkins/:/var/jenkins') {
                timeout(time: 24, unit: 'HOURS')
                {
                    cmake_build(conf)
                    dir("build"){
                        //run tests and examples
                        sh 'make -j check'
                        //we only need the ckProfiler to run the performance tests, so we pack and stash it
                        sh 'tar -zcvf ckProfiler.tar.gz bin/ckProfiler'
                        stash "ckProfiler.tar.gz"
                    }
                }
            }
        }
        return retimage
}

def Build_CK_and_Reboot(Map conf=[:]){
    try{
        Build_CK(conf)
    }
    catch(e){
        echo "throwing error exception while building CK"
        echo 'Exception occurred: ' + e.toString()
        throw e
    }
    finally{
        if (!conf.get("no_reboot", false)) {
            reboot()
        }
    }
}

500
501
502
def process_results(Map conf=[:]){
    env.HSA_ENABLE_SDMA=0
    checkout scm
503
    def image = getDockerImageName() 
504
505
506
507
508
    def prefixpath = "/opt/rocm"

    // Jenkins is complaining about the render group 
    def dockerOpts="--cap-add=SYS_PTRACE --security-opt seccomp=unconfined"
    if (conf.get("enforce_xnack_on", false)) {
509
        dockerOpts = dockerOpts + " --env HSA_XNACK=1 "
510
    }
511
    def dockerArgs = "--build-arg PREFIX=${prefixpath} --build-arg compiler_version='release' "
512
513
514
515
516
517

    def variant = env.STAGE_NAME
    def retimage

    gitStatusWrapper(credentialsId: "${status_wrapper_creds}", gitHubContext: "Jenkins - ${variant}", account: 'ROCmSoftwarePlatform', repo: 'composable_kernel') {
        try {
518
            (retimage, image) = getDockerImage(conf)
519
520
521
522
523
524
525
526
527
528
529
530
531
        }
        catch (org.jenkinsci.plugins.workflow.steps.FlowInterruptedException e){
            echo "The job was cancelled or aborted"
            throw e
        }
    }

    withDockerContainer(image: image, args: dockerOpts + ' -v=/var/jenkins/:/var/jenkins') {
        timeout(time: 1, unit: 'HOURS'){
            try{
                dir("script"){
                    if (params.RUN_FULL_QA){
                        // unstash perf files to master
532
533
534
535
536
537
538
539
540
541
542
543
                        unstash "perf_gemm.log"
                        unstash "perf_resnet50_N256.log"
                        unstash "perf_resnet50_N4.log"
                        unstash "perf_batched_gemm.log"
                        unstash "perf_grouped_gemm.log"
                        unstash "perf_conv_fwd.log"
                        unstash "perf_conv_bwd_data.log"
                        unstash "perf_gemm_bilinear.log"
                        unstash "perf_reduction.log"
                        unstash "perf_splitK_gemm.log"
                        unstash "perf_onnx_gemm.log"
                        sh "./process_qa_data.sh"
544
545
546
                    }
                    else{
                        // unstash perf files to master
547
548
549
550
                        unstash "perf_gemm.log"
                        unstash "perf_resnet50_N256.log"
                        unstash "perf_resnet50_N4.log"
                        sh "./process_perf_data.sh"
551
552
553
554
555
556
557
558
559
560
561
562
563
                    }
                }
            }
            catch(e){
                echo "throwing error exception while processing performance test results"
                echo 'Exception occurred: ' + e.toString()
                throw e
            }
        }
    }
}

//launch develop branch daily at 23:00 in FULL_QA mode
564
CRON_SETTINGS = BRANCH_NAME == "develop" ? '''0 23 * * * % RUN_FULL_QA=true''' : ""
565

JD's avatar
JD committed
566
567
pipeline {
    agent none
568
569
570
    triggers {
        parameterizedCron(CRON_SETTINGS)
    }
JD's avatar
JD committed
571
572
573
    options {
        parallelsAlwaysFailFast()
    }
574
    parameters {
575
576
577
578
        booleanParam(
            name: "BUILD_DOCKER",
            defaultValue: true,
            description: "Force building docker image (default: true)")
579
580
        string(
            name: 'COMPILER_VERSION', 
581
582
            defaultValue: 'release', 
            description: 'Specify which version of compiler to use: ck-9110, release (default), or amd-stg-open.')
583
584
585
586
        string(
            name: 'BUILD_COMPILER', 
            defaultValue: 'hipcc', 
            description: 'Specify whether to build CK with hipcc (default) or with clang.')
587
588
589
590
        booleanParam(
            name: "RUN_FULL_QA",
            defaultValue: false,
            description: "Select whether to run small set of performance tests (default) or full QA")
591
592
593
594
        booleanParam(
            name: "TEST_NODE_PERFORMANCE",
            defaultValue: false,
            description: "Test the node GPU performance (default: false)")
595
596
597
598
599
600
601
602
    }
    environment{
        dbuser = "${dbuser}"
        dbpassword = "${dbpassword}"
        dbsship = "${dbsship}"
        dbsshport = "${dbsshport}"
        dbsshuser = "${dbsshuser}"
        dbsshpassword = "${dbsshpassword}"
603
        status_wrapper_creds = "${status_wrapper_creds}"
604
605
        gerrit_cred="${gerrit_cred}"
        DOCKER_BUILDKIT = "1"
606
    }
JD's avatar
JD committed
607
    stages{
608
609
610
611
612
613
614
615
616
617
618
619
620
        stage("Build Docker"){
            when {
                expression { params.BUILD_DOCKER.toBoolean() }
            }
            parallel{
                stage('Docker /opt/rocm'){
                    agent{ label rocmnode("nogpu") }
                    steps{
                        buildDocker('/opt/rocm')
                    }
                }
            }
        }
JD's avatar
JD committed
621
        stage("Static checks") {
622
623
624
625
            when {
                beforeAgent true
                expression { !params.TEST_NODE_PERFORMANCE.toBoolean() }
            }
JD's avatar
JD committed
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
            parallel{
                // enable after we move from hipcc to hip-clang
                // stage('Tidy') {
                //     agent{ label rocmnode("nogpu") }
                //     environment{
                //         // setup_cmd = "CXX='/opt/rocm/bin/hipcc' cmake -DBUILD_DEV=On .. "
                //         build_cmd = "make -j\$(nproc) -k analyze"
                //     }
                //     steps{
                //         buildHipClangJobAndReboot(build_cmd: build_cmd, no_reboot:true, prefixpath: '/opt/rocm', build_type: 'debug')
                //     }
                // }
                stage('Clang Format') {
                    agent{ label rocmnode("nogpu") }
                    environment{
JD's avatar
JD committed
641
                        execute_cmd = "find .. -iname \'*.h\' \
JD's avatar
JD committed
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
                                -o -iname \'*.hpp\' \
                                -o -iname \'*.cpp\' \
                                -o -iname \'*.h.in\' \
                                -o -iname \'*.hpp.in\' \
                                -o -iname \'*.cpp.in\' \
                                -o -iname \'*.cl\' \
                                | grep -v 'build/' \
                                | xargs -n 1 -P 1 -I{} -t sh -c \'clang-format-10 -style=file {} | diff - {}\'"
                    }
                    steps{
                        buildHipClangJobAndReboot(setup_cmd: "", build_cmd: "", execute_cmd: execute_cmd, no_reboot:true)
                    }
                }
            }
        }
657
658
    
		stage("Build CK and run Tests")
659
660
661
        {
            parallel
            {
662
                stage("Build CK and run Tests")
663
                {
664
                    agent{ label rocmnode("gfx908 || gfx90a") }
665
                    environment{
666
667
                        setup_args = "${params.COMPILER_VERSION == "ck-9110" ? """ -DBUILD_DEV=Off -DCMAKE_INSTALL_PREFIX=../install -DGPU_TARGETS="gfx908;gfx90a" -DCMAKE_CXX_FLAGS="-O3 -Xclang -mlink-builtin-bitcode -Xclang /opt/rocm/amdgcn/bitcode/oclc_abi_version_400.bc" """ : """ -DBUILD_DEV=Off -DCMAKE_INSTALL_PREFIX=../install -DGPU_TARGETS="gfx908;gfx90a" -DCMAKE_CXX_FLAGS="-O3 " """ }"
                        execute_args = "${params.COMPILER_VERSION == "ck-9110" ? """ cd ../client_example && rm -rf build && mkdir build && cd build && cmake -D CMAKE_PREFIX_PATH="${env.WORKSPACE}/install;/opt/rocm" -DGPU_TARGETS="gfx908;gfx90a" -DCMAKE_CXX_FLAGS="-O3 -Xclang -mlink-builtin-bitcode -Xclang /opt/rocm/amdgcn/bitcode/oclc_abi_version_400.bc" -D CMAKE_CXX_COMPILER="${build_compiler()}" .. && make -j """ : """ cd ../client_example && rm -rf build && mkdir build && cd build && cmake -D CMAKE_PREFIX_PATH="${env.WORKSPACE}/install;/opt/rocm" -DGPU_TARGETS="gfx908,gfx90a" -DCMAKE_CXX_FLAGS="-O3" -D CMAKE_CXX_COMPILER="${build_compiler()}" .. && make -j """ }"
668
669
                    }
                    steps{
670
                        Build_CK_and_Reboot(setup_args: setup_args, config_targets: "install", no_reboot:true, build_type: 'Release', execute_cmd: execute_args, prefixpath: '/usr/local')
JD's avatar
JD committed
671
672
                    }
                }
673
674
            }
        }
675
676
677
678
679

 		/*
        //at present this stage only builds binaries. 
        //we will now build all binaries in a separate stage.
        //once we have some tests to run in this stage, we can enable it again.
Chao Liu's avatar
Chao Liu committed
680
681
        stage("Client App")
        {
682
683
684
685
            when {
                beforeAgent true
                expression { !params.TEST_NODE_PERFORMANCE.toBoolean() }
            }
Chao Liu's avatar
Chao Liu committed
686
687
688
689
690
691
            parallel
            {
                stage("Run Client App")
                {
                    agent{ label rocmnode("gfx908")}
                    environment{
692
693
                        setup_args = "${params.COMPILER_VERSION == "ck-9110" ? """ -DBUILD_DEV=Off -DCMAKE_INSTALL_PREFIX=../install -DGPU_TARGETS="gfx908;gfx90a" -DCMAKE_CXX_FLAGS="-O3 -Xclang -mlink-builtin-bitcode -Xclang /opt/rocm/amdgcn/bitcode/oclc_abi_version_400.bc" """ : """ -DBUILD_DEV=Off -DCMAKE_INSTALL_PREFIX=../install -DGPU_TARGETS="gfx908;gfx90a" -DCMAKE_CXX_FLAGS="-O3 " """ }"
                        execute_args = "${params.COMPILER_VERSION == "ck-9110" ? """ cd ../client_example && rm -rf build && mkdir build && cd build && cmake -D CMAKE_PREFIX_PATH="${env.WORKSPACE}/install;/opt/rocm" -DGPU_TARGETS="gfx908;gfx90a" -DCMAKE_CXX_FLAGS="-O3 -Xclang -mlink-builtin-bitcode -Xclang /opt/rocm/amdgcn/bitcode/oclc_abi_version_400.bc" -D CMAKE_CXX_COMPILER="${build_compiler()}" .. && make -j """ : """ cd ../client_example && rm -rf build && mkdir build && cd build && cmake -D CMAKE_PREFIX_PATH="${env.WORKSPACE}/install;/opt/rocm" -DGPU_TARGETS="gfx908;gfx90a" -DCMAKE_CXX_FLAGS="-O3" -D CMAKE_CXX_COMPILER="${build_compiler()}" .. && make -j """ }"
Chao Liu's avatar
Chao Liu committed
694
695
696
697
698
699
700
                    }
                    steps{
                        buildHipClangJobAndReboot(setup_args: setup_args, config_targets: "install", no_reboot:true, build_type: 'Release', execute_cmd: execute_args, prefixpath: '/usr/local')
                    }
                }
            }
        }
701
        */
702
703
704
705
        stage("Performance Tests")
        {
            parallel
            {
706
                stage("Run ckProfiler: gfx908 or gfx90a")
707
                {
708
709
                    when {
                        beforeAgent true
710
                        expression { !params.RUN_FULL_QA.toBoolean() && !params.TEST_NODE_PERFORMANCE.toBoolean() }
711
                    }
712
                    options { retry(2) }
713
                    agent{ label rocmnode("gfx908 || gfx90a")}
714
                    environment{
715
                        setup_args = "${params.COMPILER_VERSION == "ck-9110" ? """ -DGPU_TARGETS="gfx908;gfx90a" -DCMAKE_CXX_FLAGS=" -O3 -Xclang -mlink-builtin-bitcode -Xclang /opt/rocm/amdgcn/bitcode/oclc_abi_version_400.bc" -DBUILD_DEV=On """ : """ -DGPU_TARGETS="gfx908;gfx90a" -DCMAKE_CXX_FLAGS=" -O3 " -DBUILD_DEV=On """}"
716
                   }
717
                    steps{
718
                        runPerfTest(setup_args:setup_args, config_targets: "ckProfiler", no_reboot:true, build_type: 'Release')
719
720
721
722
                    }
                }
                stage("Run ckProfiler: gfx90a")
                {
723
724
                    when {
                        beforeAgent true
725
                        expression { params.RUN_FULL_QA.toBoolean() || params.TEST_NODE_PERFORMANCE.toBoolean() }
726
                    }
727
                    options { retry(2) }
728
729
                    agent{ label rocmnode("gfx90a")}
                    environment{
730
                        setup_args = "${params.COMPILER_VERSION == "ck-9110" ? """ -DGPU_TARGETS="gfx90a" -DCMAKE_CXX_FLAGS=" -O3 -Xclang -mlink-builtin-bitcode -Xclang /opt/rocm/amdgcn/bitcode/oclc_abi_version_400.bc" -DBUILD_DEV=On """ : """ -DGPU_TARGETS="gfx90a" -DCMAKE_CXX_FLAGS=" -O3 " -DBUILD_DEV=On """}"
731
                    }
732
                    steps{
733
                        runPerfTest(setup_args:setup_args, config_targets: "ckProfiler", no_reboot:true, build_type: 'Release')
734
735
736
737
                    }
                }
            }
        }
738
739
740
741
        stage("Process Performance Test Results")
        {
            parallel
            {
742
                stage("Process results"){
743
744
                    agent { label 'mici' }
                    steps{
745
                        process_results()
746
747
748
749
750
                    }
                }
            }
        }

751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
        /* enable after the cmake file supports packaging
        stage("Packages") {
            when {
                expression { params.BUILD_PACKAGES && params.TARGET_NOGPU && params.DATATYPE_NA }
            }
            parallel {
                stage("Package /opt/rocm") {
                    agent{ label rocmnode("nogpu") }
                    steps{
                        buildHipClangJobAndReboot( package_build: "true", prefixpath: '/opt/rocm', gpu_arch: "gfx906;gfx908;gfx90a")
                    }
                }
            }
        }
        */
JD's avatar
JD committed
766
    }
767
}