Compare commits

...

89 Commits

Author SHA1 Message Date
Alex Xu
deb4895b11 Merge branch 'develop' into roc-6.3.x 2024-12-20 18:42:53 -05:00
Peter Park
47502421e1 fix merge conflicts (#4185) 2024-12-20 18:34:17 -05:00
Peter Park
bf9727db74 Fix links in docs (#4182)
* fix links

* missing space

* fix link in known issue

* Remove programming guide from TOC
2024-12-20 17:21:30 -05:00
Peter Park
49d253dd13 Fix broken links in 6.3.1 docs (#4180)
* fix links

* missing space
2024-12-20 16:51:39 -05:00
alexxu-amd
94f97d4428 Merge pull request #4179 from ROCm/amd/alexxu12/update631
Change version from 6.3.0 to 6.3.1
2024-12-20 16:18:00 -05:00
Pratik Basyal
33891d6fd0 Bar memory update develop (#4168)
* Bar Memory page added

* Leo's feedback incorporated

* Spell check fixed

* SME review feedback incorporated

* Feedback updated

* Indentation fixed

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-20 16:05:49 -05:00
Alex Xu
859923e595 change version from 6.3.0 to 6.3.1 2024-12-20 15:53:46 -05:00
alexxu-amd
74f3e26b62 Merge pull request #4177 from ROCm/sync-develop-from-internal
Sync develop from internal
2024-12-20 15:18:42 -05:00
Alex Xu
09a5a2f23a resolve merge conflict 2024-12-20 15:06:43 -05:00
Peter Park
59f217be22 fix merge conflicts in workload.rst 2024-12-20 14:52:30 -05:00
Alex Xu
2fd18ab1b6 revert KFD change back 2024-12-20 14:23:57 -05:00
Alex Xu
d275733631 Merge remote-tracking branch 'internal/develop' into sync-develop-from-internal 2024-12-20 14:00:20 -05:00
darren-amd
16e18bf5e6 External CI: Add rocprof-sdk and aqlprofile dependencies to rocprof-sys (#4176) 2024-12-20 13:59:39 -05:00
Joseph Macaranas
c9dd86f2d2 External CI: Add toggle for torchvision build and test (#4173)
- Recent vision compilation has been failing, and debugging hasn't been fruitful in finding cause.
- Should unblock nightly job to at least build and test pytorch while debug effort continues after the holidays.
- pytorch build and test is unblocked by temporarily patching the composable_kernel submodule on upstream pytorch to latest develop, until that submodule is updated to have explicit cast for hneg.
2024-12-20 13:59:04 -05:00
Peter Park
95845105a5 Fix 6.3.1 links (#269) 2024-12-20 13:23:26 -05:00
Pratik Basyal
14a6fd5837 Release.md and autotag template updated for 6.3.1 release prep (#268)
* Release ready final changes and template update

* wordlist updated

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-20 12:56:13 -05:00
Pratik Basyal
eb06305a88 Release management feedback incorporated (#267)
* Release management feedback incorporated

* RN updated

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-20 11:59:50 -05:00
Pratik Basyal
771518a2aa ROCm runfile link reverted to internal (#266)
Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-20 10:00:45 -05:00
Pratik Basyal
3e714f683c Known issue draft added (#265)
* Known issue draft added

* SME feedback incorporated

* Known issue updated

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-20 06:57:41 -05:00
Pratik Basyal
1998affd4f Link update for install instruction (#262)
* Link update for install instruction

* Debian related changes

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-20 06:47:22 -05:00
Peter Park
8d3de707e2 add MI325 to GPU specs page (#264) 2024-12-19 15:20:00 -05:00
randyh62
5f86dba37c Update RELEASE.md (#263)
remove HIP 6.3.1 content that is not in the release
2024-12-19 11:34:11 -08:00
Pratik Basyal
9a602592a0 Mi325X GPU support (#260)
* MI325X support added

* RN updated

* MI325x moved to OS hardware support

* Update RELEASE.md

Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com>

* Update RELEASE.md

* Update RELEASE.md

Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com>

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com>
2024-12-19 12:46:05 -05:00
Pratik Basyal
6d524affd2 Doc highlight updated (#259)
Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-19 10:57:58 -05:00
alexxu-amd
988da8a0d1 Merge pull request #256 from ROCm/sync-develop-from-external
Sync develop from external
2024-12-19 09:52:54 -05:00
alexxu-amd
758e8a33db Merge branch 'develop' into sync-develop-from-external 2024-12-19 09:48:30 -05:00
Istvan Kiss
21d26e52d0 Add graph safe support 2024-12-19 14:48:58 +01:00
Pratik Basyal
134ee09631 Footnote updated (#258)
Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-18 18:43:47 -05:00
randyh62
4f10f22920 Rocm image 631 (#257)
* Add files via upload

Add Debian OS without TransferBench

* Delete docs/data/rocm-software-stack-6_3_0.jpg

remove 6-3-0 image

* Update what-is-rocm.rst

Remove link and description for TransferBench

* Update rocm-tools.md

remove TransferBench

* Update what-is-rocm.rst

update ROCm Software Stack image name

* Add files via upload

Add correct image
2024-12-18 15:12:33 -08:00
Alex Xu
935a55f703 resolve more merge conflicts 2024-12-18 16:49:16 -05:00
Alex Xu
45d59ffdba add back overwritten PR 2024-12-18 16:41:24 -05:00
alexxu-amd
c2be7ee900 Merge branch 'develop' into sync-develop-from-external 2024-12-18 16:19:12 -05:00
Pratik Basyal
279a241c11 Quickupdates release631 develop (#255)
* Transferbench added

* Minor fix

* Table alignment fixed

* Review feedback

* Leo's feedback incorporated

* Fixed issue added

* Compatibility Matrix table fixed

* JAX version updated

* Debian support added

* Transferbench added

* Debian footnote updated

* Debian added to wordlist

* Debian footnote updated

* wordlist updated

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-18 16:08:13 -05:00
Alex Xu
0356ffd148 Merge remote-tracking branch 'external/develop' into sync-develop-from-external 2024-12-18 15:57:08 -05:00
randyh62
cf3c4d1e67 Whatis 631 (#253)
* Update what-is-rocm.rst

minor change

* Update rocm-tools.md

minor change
2024-12-18 07:11:26 -08:00
Pratik Basyal
cbe105ae8f TransferBench Release Notes Updated (#252)
* Transferbench added

* Minor fix

* Table alignment fixed

* Review feedback

* Leo's feedback incorporated

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-18 00:00:51 -05:00
randyh62
5c5b5cce73 Add files via upload (#250)
Add ROCm image with TransferBench
2024-12-17 15:37:09 -08:00
Pratik Basyal
72dce7937b Add TransferBench to release notes for 6.3.1 (#251)
* Transferbench added

* Minor fix

* Table alignment fixed

* Review feedback

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-17 17:49:07 -05:00
randyh62
08d9286cd5 Update ROCm software stack image for 6.3.1 (#245) 2024-12-17 11:55:51 -08:00
Pratik Basyal
6a7d8654ad Revamped PCIe into new format and incorporated style guide (#4051)
* Revamped PCIe into new format and incorporated style guide

* Title case fixed

* Quick fix and changes

* Added RMW to wordlist and updated titles

* Grammatical fixes incorporated

* Sandra's review feedback incorporated

* Removed PCIe3 feature reference

* Leo's feedback incorporated

* Sandra's feedback incorporated

* Replaced execute with run

* Replaced executing with running

* SME review feedback incorporated

* Minor feedback updated

* Sandra's feedback incorporated

* Filename renamed

* File rename changes updated

* Document title updated

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-17 12:00:00 -05:00
Peter Park
c5ee1196c4 6.3.1 release notes style update (#249)
* update RELEASE.md

(cherry picked from commit ad449acd31c57c93be0d9f7dc8270f48747c11e1)

* Update RELEASE.md

Co-authored-by: Pratik Basyal <pratik.basyal@amd.com>

---------

Co-authored-by: Pratik Basyal <pratik.basyal@amd.com>
2024-12-16 15:18:01 -05:00
darren-amd
dc648ad764 External CI: Disable rocprof-system openmp target example (#4166) 2024-12-16 14:38:55 -05:00
Peter Park
f9dbc1f21f add megatron training doc (#4159)
* add megatron training doc

update toc

add images

update formatting and wording

formatting

update formatting

update conf.py

update formatting

update docker img

tweak formatting

Fix stuff

fix mock-data/data-path

add specific commit hash to checkout

update docker pull tag

fix docker run cmd and examples path

fix docker cmd

* wording

words

words

* improve title
2024-12-16 13:37:35 -05:00
Yanyao Wang
a857597340 Merge pull request #4162 from WBobby/develop-pr
Update build scripts of ROCm6.3 release to develop branch
2024-12-16 12:23:27 -06:00
Yanyao Wang
8c036531e8 Merge pull request #4163 from WBobby/roc-6.3.x-pr
Update build scripts of ROCm6.3 release to roc-6.3.x branch
2024-12-16 12:23:11 -06:00
Pratik Basyal
e839c63a01 Updated stagin reference for review (#247)
Co-authored-by: prbasyal <prbasyal@amd.com>
2024-12-16 13:07:42 -05:00
Pratik Basyal
6d997135a5 Update Compatibility Matrix for 6.3.1 (#240)
* Updated for 6.3.1

* Compatible version updated from RC1 build

Co-authored-by: Peter Park <peter.park@amd.com>

* Comptibility table and rst updated

* Compatible version updated from RC1 build

Co-authored-by: Peter Park <peter.park@amd.com>

* Peter's review feedback incorporated

Co-authored-by: Peter Park <peter.park@amd.com>

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
Co-authored-by: Peter Park <peter.park@amd.com>
2024-12-16 11:19:01 -05:00
Pratik Basyal
2d6a601253 Update to 6.3.1 Release Notes (#236)
* 6.3.1 Release notes (#224)


* New Release highlight on offline installer added. OS change, Known Issues, Resolved issues, and upcoming changes copied from 6.3.0 and updated version

---------

Co-authored-by: prbasyal <prbasyal@amd.com>

* Updates to release notes (#229)

* Updates to release notes

* & -> and

* Updated the component changes, table, release highlights, and fixed i… (#232)

* Updated the component changes, table, release highlights, and fixed issues

* Version number and heading title fixed

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Version transition updated

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
Co-authored-by: Peter Park <peter.park@amd.com>
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Rn 631 custombranch (#234)

* Updated the component changes, table, release highlights, and fixed issues

* Version number and heading title fixed

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Version transition updated

* OS and Hardware compatibility updated

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
Co-authored-by: Peter Park <peter.park@amd.com>
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* [6.3.1 release notes] Add rocprof-sys changes to RN (#235)

* remove extra sections

* add rocprof-sys changelog

* add omni fixed issues and ami smi cl

* Update RELEASE.md

Version transition added in the table

* add documentation update note

* Broken link fixed

---------

Co-authored-by: Pratik Basyal <pratik.basyal@amd.com>

* added edited version of the migraphx changelog and removed CK entry (#238)

Co-authored-by: Pratik Basyal <pratik.basyal@amd.com>

* Updating ROCm-internal with 6.3.1 release notes changes (#241)

* Updated date and version

* Typos and wording fixed

* Minor fix

* Documentation update added

* MIGraphX change dropped

* Debian support removed

* New release highlight added

* HIPRand version changed

* Cross-reference to Per queue added

* Leo's review feedback incorporated

* HIP optimized section updated

* Istinct and Peter's feedback added

---------

Co-authored-by: prbasyal <prbasyal@amd.com>

* Fix changelog and new documentation note (#246)

* fix amd smi and add training a model using megatron note

* update workload tuning doc note

* fmt

---------

Co-authored-by: prbasyal <prbasyal@amd.com>
Co-authored-by: spolifroni-amd <Sandra.Polifroni@amd.com>
Co-authored-by: Peter Park <peter.park@amd.com>
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>
2024-12-16 09:30:20 -05:00
Wang, Yanyao
484cbefc2e Update build scripts of ROCm6.3 release to roc-6.3.x branch 2024-12-15 17:35:58 -08:00
Wang, Yanyao
a2d128749c Update build scripts of ROCm6.3 release to develop branch 2024-12-15 17:06:23 -08:00
Joseph Macaranas
bacb49681e External CI: Typo in rocPyDecode Pipeline Parameter (#4157) 2024-12-13 14:35:35 -05:00
alexxu-amd
721b60d52f Merge pull request #4155 from amd-jnovotny/user-kernel-space-rocm-roc63x
Cherry-pick to roc-6.3.x: Change reference to kernel-mode GPU compute driver in ROCm (#4147)
2024-12-13 13:15:06 -05:00
Jeffrey Novotny
8ebe7be283 Change reference to kernel-mode GPU compute driver in ROCm (#4147)
* Change reference to kernel-mode GPU compute driver in ROCm

* More changes for kernel-mode terminology

* Fix linting

(cherry picked from commit 04fdc08328)
2024-12-13 12:13:15 -05:00
Jeffrey Novotny
04fdc08328 Change reference to kernel-mode GPU compute driver in ROCm (#4147)
* Change reference to kernel-mode GPU compute driver in ROCm

* More changes for kernel-mode terminology

* Fix linting
2024-12-13 11:46:02 -05:00
darren-amd
1b33f1d7da External CI: llvm project comgr disable spirv (#4154)
External CI: add flag to disable SPIRV from comgr build in llvm-project
2024-12-13 10:51:36 -05:00
Joseph Macaranas
fd067f7b3b External CI: HIP shared library symlinks for rocPyDecode (#4153)
- Modifying the HIP shared libraries installed to follow the Linux symbolic link convention resolves test failures in rocPyDecode.
2024-12-13 09:58:25 -05:00
spolifroni-amd
2a7520f08a Added MIGraphX changes (#4150)
* Added MIGraphX changes

* removed gfx support

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

---------

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>
2024-12-12 11:19:28 -05:00
randyh62
a591218531 ROCm software stack update for 6.3.1 (#242) 2024-12-11 15:29:41 -08:00
Joseph Macaranas
5271c2c82d External CI: MIOpen Test Parameters (#4148)
- Exclude lone, consistently failing MIOpen test.
- test_rnn_seq_api is the only ctest failure, so let's filter it out for now to easily identify new failures.
2024-12-11 13:31:12 -05:00
JeniferC99
59a928f3a7 Merge pull request #4137 from ROCm/JeniferC99-patch-1
Update default.xml
2024-12-10 11:47:38 -06:00
David Galiffi
22572a9857 Add TransferBench and hipBLAS-common 2024-12-09 18:37:20 -05:00
randyh62
49e50b93c6 Update index.md (#4144) (#4146)
Remove Programming Guide topic from "How to"
2024-12-09 12:17:54 -08:00
Istvan Kiss
3354099b9c Remove GPU memory page 2024-12-09 17:23:57 +01:00
David Galiffi
794b34f40e Update default.xml
Fixed merge conflicts
2024-12-09 11:18:39 -05:00
David Galiffi
25ef417b31 Merge branch 'develop' into JeniferC99-patch-1 2024-12-09 11:17:17 -05:00
Peter Park
78f9adc6ec fix rccl hip streams section in workload tuning guide (#4140) 2024-12-09 11:06:12 -05:00
David Galiffi
4abcae54a8 Update default.xml (#4136)
Add rocJPEG
Rename omniperf to rocprofiler-compute
Rename omnitrace to rocprofiler-compute
2024-12-09 10:56:07 -05:00
JeniferC99
2690506e64 Update default.xml
SWDEV-502858
Rename  Omnitrace and Omniperf
2024-12-06 23:01:56 -06:00
darren-amd
3dffe1998a External CI: add aomp dependency to rocprofiler-sdk (#4135)
External CI: Add aomp dependency for rocprofiler-sdk
2024-12-06 16:20:09 -05:00
Peter Park
059c2cd9a4 6.3.0 release notes (#199)
* generate 6.3.0 RELEASE.md

* add 6.3.0 os/hw support

* regenerate changelog

* update table

* add amd smi and fix fmt

* add rocjpeg note

* add missed changelog entries

* update ga date

* add SHARK toolkit introduced note

update SHARK note

* Edited some components (#202)

* Edited some components

* fixed formatting on rocal

* markdown fail on the last commit; fixed

* capitalization fix

* Copy edit component change logs (#203)

* fix some formatting

* fix table and add OpenCL note

fix fmt

fix more formatting

* add radeon note

* add rocmsmi

* Updated hipCUB, rocPrim, and rocThrust (#206)

* fix some stuff

* add transferbench

* Edits to RCCL 6.3 change log (#207)

* Update tools/autotag/templates/upcoming_changes/6.3.0.md

* fix formatting

* fix sphinx underline warning

* add @lpaoletti's highlights

* fix os support

* add missing kernel version

* fix heading

* add bitsandbytes ki

* Copy edits to release notes (#208)

* Copy edits to release notes

* Additional updates to release notes

* updated shark AI toolkit description

* fix formatting

* update opencl

* update opencl

fixes and updates

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* fix omnitools rename text

* Apply suggestions from code review

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

* Update RELEASE.md

* Update RELEASE.md

* Update RELEASE.md

* Update RELEASE.md

* Update RELEASE.md

* update omniperf and tesile notes

* Update RELEASE.md

* Update RELEASE.md

* Update RELEASE.md

* Update RELEASE.md

* Update RELEASE.md

* Apply suggestions from code review

* Apply suggestions from code review

* Apply suggestions from code review

* Apply suggestions from code review

* made some copy edits (#209)

* Apply suggestions from code review

* Update RELEASE.md

* Apply suggestions from code review

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* indent

* add more highlights

* update shark urls

* add omni notes

* Apply suggestions from code review

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* update some changelogs

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* update some cls

* and missed changelogs

* add missed component updates

* fix links

* add amdgpu-dkms highlight

* Update RELEASE.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* change links

* add fixed issues

* @neon60's changes

Co-authored-by: Istvan Kiss <neon60@gmail.com>

* Apply suggestions from code review

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>
Co-authored-by: Swati Rawat <120587655+SwRaw@users.noreply.github.com>

* rm extra hip docs

* add hip links

* add fixed issue

fix

* Update RELEASE.md

Co-authored-by: Istvan Kiss <neon60@gmail.com>

* Update RELEASE.md

Co-authored-by: Istvan Kiss <neon60@gmail.com>

* Update RELEASE.md

Co-authored-by: Istvan Kiss <neon60@gmail.com>

* fix ri

* fix zebra

* Update RELEASE.md

Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com>

* rm extra amd smi info

* Apply suggestions from code review

Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com>
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* add more about omni renmae

fix rename stuff

* Update RELEASE.md

Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com>

* Update RELEASE.md

Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com>

* fix formatting

* wording

* fix link

* update aotriton

* remove libraries performance improved

* fix rhel version

* fix urls

shorten title

* Apply suggestions from code review

Co-authored-by: Swati Rawat <120587655+SwRaw@users.noreply.github.com>

* Release notes updates (#212)

* Made language more precise (#211)

MIVisionX and rocAL were changed. An awkward sentence in rocAL was also fixed.

* add rocprofiler

* add rdc

add rdc entry

* Update RELEASE.md

Co-authored-by: Istvan Kiss <neon60@gmail.com>

* Update RELEASE.md

Co-authored-by: Istvan Kiss <neon60@gmail.com>

* Update RELEASE.md

Co-authored-by: Swati Rawat <120587655+SwRaw@users.noreply.github.com>

* remove bitsandbytes known issue

* fix missed hip doc

* update rocprof-compute version to 3.0.0

* remove words

* change hiprand ver to 2.11.0

* update new components descriptions

* add #

* fix tensile versions

* fix versions and add missed cls

* Update RELEASE.md

Co-authored-by: Istvan Kiss <neon60@gmail.com>

* remove resolved issue for #3493

* add rdc note

* add hiprand known issue

add hiprand known issue

add asterisk for hiprand ki

asterisk formatting

asterisk

link asterisk

* rdc known issue

* @lpaoletti updates

* @wenchenvincent add CK to Transformer Engine note

* fix links

fix links

* add roct thunk interface note

* rm 'previously'

* Apply suggestions from code review

Co-authored-by: Istvan Kiss <neon60@gmail.com>

* add known issues

* add mi300x cpfw known issue

* add mi300x cpfw known issue

add note

* spacing

* update te error KI

* rm incorrect user impact in TE known issue

* correct description of transformer engine fatal python error known issue

* update autotag/templates

* fix order

* fix typo

* update .wordlist.txt w/ lib names

* add missing css classes

* remove ROCT-Thunk-Interface from ROCm licenses

* add rocJPEG LICENSE

* fix table zebra b/c added rows

* fix capitalization in toc

* update URLs post-review

* update AMD SMI changelog

* update ROCm SMI changelog

* add opencl icd stale file kI

words

* remove Azure Linux

* update omnitrace note

* add mi200 DLM known issue

* update omnitrace note

update omnitrace note

wording

update omnitrace note

* update 6.3 ga to 11/26

* update KIs wording

* Update tools/autotag/templates/highlights/6.3.0.md

Co-authored-by: Istvan Kiss <neon60@gmail.com>

* Update tools/autotag/templates/highlights/6.3.0.md

Co-authored-by: Istvan Kiss <neon60@gmail.com>

* update TransferBench note

* remove transferbench

remove transferbench

* remove gfx12, 1151

* remove sr-iov

* rm tb

* css classes

* rm gfx12

* add back transferbench

* add transferbench to table

* rm transferbench, add as KI

* update transferbench KI workaround

* add rocprof-comp KI

fix

* fix tensile

* add backward weights conv KI

update

* remove RHEL 8.9 from OS EOS

* remove mi200 perf drop for DLMs

* add RHEL 8.9 to end of support OSes

* add omniperf/omnitrace KIs

* remove bf16 statement in mi300x KI

* update rvs versions in compat

* add amd smi KI

update

update

* words

* update GA date for 6.3.0

* add rvs KI

* add KI links

same

* rvs in compat

* update tf versions

* add rvs changelog

* update rn templates

* add possessives to wordlist

---------

Co-authored-by: spolifroni-amd <Sandra.Polifroni@amd.com>
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>
Co-authored-by: randyh62 <42045079+randyh62@users.noreply.github.com>
Co-authored-by: Istvan Kiss <neon60@gmail.com>
Co-authored-by: Swati Rawat <120587655+SwRaw@users.noreply.github.com>
Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com>
2024-12-03 15:16:38 -05:00
Peter Park
f53faa19ea update compatibility card to install in index.md (#221) 2024-12-02 16:32:18 -05:00
Peter Park
c5bf6c39ca remove RHEL 8.9 from compat (#223) 2024-12-02 10:57:52 -05:00
Peter Park
e677ebcb47 Trim TOC sections covered in Compatibility Matrix (#219)
* trim toc

* better org
2024-11-27 12:32:21 -05:00
Peter Park
a52c98cb2b 6.3.0 compat matrix WIP (#197)
* add tentative historical compat matrix

* update historical compat

* update compat

* update rocprof-compute version to 3.0.0

* update historical compat matrix

* update os support in historical

* update compat and historical

* fix compat formatting

* fix footnotes

* update kernel versions

* remove mi300 footnote for gfx942

* rm extra kernel ver

* add azure linux

* update tested user space vers

* spellcheck

* add roct to wordlist

* remove azure linux

* update thrust/cub versions

* fix

* rocr

rocr

* extra underscore

* add back RHEL 8.9
2024-11-27 12:00:55 -05:00
Alex Xu
d4ad0a838d update rocm-docs-core from 1.8.3 to 1.9.0 2024-11-22 19:20:15 -05:00
randyh62
e9e9fa4ba5 Change Common Language Runtime to Compute Language Runtime (#200)
* Change Common Language Runtime to Compute Language Runtime

* change rocJPEG description

* update ROCm Software Stack image
2024-11-22 15:45:26 -08:00
randyh62
c18694b0fe add azure image (#214) 2024-11-22 08:04:01 -08:00
randyh62
d83ed9d58a Update How to TOC and Index (#210) 2024-11-18 10:19:32 -08:00
alexxu-amd
0a237dfd42 Sync develop from external repo (#205)
* Update version list with 6.2.0 (#3505) (#3506)

* Fix link to meta-llama finetuning recipes

* Spellcheck fixes in release notes templates (#3526) (#3548)

* fix spelling in 5.4.x templates

* add to wordlist

* update templates

update wordlist

* remove extra_components

rm extra_components

* fix spelling

Co-authored-by: Peter Park <peter.park@amd.com>

* Fix link to rocr debug agent (#3533)

Co-authored-by: Sam Wu <22262939+samjwu@users.noreply.github.com>

* Fix intersphinx links (#3546)

* update fw install links

* fix more intersphinx links

* fix more links

* add rocPyDecode repo to ROCm6.2 manifest file (#3541) (#3553)

Co-authored-by: Yanyao Wang <yanywang@amd.com>
Co-authored-by: Wang, Yanyao <yanyao.wang@amd.com>

* Fix typo for TFLOPs metric in MI250 architecture page

* Add rocm-examples to default.xml (#3583)

* Add rocm 6.2.0 manifest file for rocm-build scripts (#3538)

* Add rocm 6.2.0 manifest file for rocm-build scripts

Signed-off-by: David Galiffi <David.Galiffi@amd.com>

* Add "rocm-examples"

---------

Signed-off-by: David Galiffi <David.Galiffi@amd.com>

* Add a section on increasing memory allocation to the MI300A system op… (#3587)

* Add a section on increasing memory allocation to the MI300A system optimization guide

* Addition to wordlist

* Change GB to GiB for consistency

* Standardize GiB/KiB spacing

* Minor wording changes

* Update build scripts for ROCm6.2 release

* fix README.md for Ubuntu24 docker

* Correct ttm to amdttm (#3648)

* Expand the section on changing thread affinity (#3653)

* Expand the section on changing thread affinity

* Clarify the methods for configuring allocatable memory settings

* Small correction

* Update model-quantization.rst to import `BitsAndBytesConfig` from transformers library (#3638)

* remove unneeded file (#3663)

* Fix intersphinx links (#3668)

* fix links in install.rst

* fix links in sys opt guides

* Add introduction and links to the new guide to the vLLM optimized Doc… (#3637)

* Add introduction and links to the new guide to the vLLM optimized Docker image on AMD Infinity Hub

* Update target link for the Docker vLLM guide

* Change target URL

* Change link target URL again

* Fixed broken link to RISC-V documentation

* Add FBGEMM/FBGEMM_GPU to the Model acceleration libraries page (#3659)

* Add FBGEMM/FBGEMM_GPU to the Model acceleration libraries page

* Add words to wordlist and fix a typo

* Add new sections for Docker and testing

* Incorporate comments from the external review

* Some minor edits and clarifications

* Incorporate further review coments and fix test section

* Add comment to test section

* Change git clone command for FBGEMM repo

* Change Docker command

* Changes from internal review

* Fix linting issue

* Fixed broken links for tensile, rocprofiler, roctracer, hipify, rocm-cmake

* add missing make command to bitsandbytes install commands (#3722)

* Update link to rocRAND data type support (#3736)

* Fix Radeon link and point at R6.1.3 as absolute link (#3757)

* Fix Radeon link and point at R6.1.3 as absolute link (#3757)

* Include rocal version change in the highlights (#177)

* Include rocal version change in the highlights

* Reworded rocal known issues and added link to rocal in highlights

* Update ROCm manifest to 6.2.1

* Update ROCm branch name

* Add 6.2.1 to version list (#3770)

* Add links to GH issues in 6.2.1 release notes (#3769)

* add MAD page

* link to GitHub issues in release notes known issues

* update templates for 6.2.1

* Revert "add MAD page"

This reverts commit 9cce72bba3.

* update wordlist for spellcheck linter

* add rccl note

* update rocal version change heading to be more obvious

* make rocal note more specific

* fix missing space

* fix capitalization

* Update RCCL known issue wording (#3775)

* add MAD page

* fix wording in RCCL known issue

* Revert "add MAD page"

This reverts commit c81d0f3b0a.

* update llvm version for 6.2.1 (#3779)

* Fix broken links in 6.2.1 release notes (#3782)

* External CI: Replace libomp dependencies with aomp (#3781)

Add roctracer dependency for hipBLAS and rocWMMA testing

* External CI: Add rocprofiler v1 and v2 smoke tests (#3784)

* External CI: ROCgdb smoke tests (#3785)

- Since this is an autotools project and not cmake, build and test on gfx942 system instead of separating into two jobs. Pipeline time is short anyway.
- Follow build instructions to update build flags and to incorporate the ROCdbgapi.
- Results are not parsed and graphed, but the log contents are printed at the end. This was helpful for debugging and will be kept in the pipeline, as the make check-gdb command's output was not helpful on its own.

* External CI: rocPyDecode Smoke Test (#3786)

* External CI: omniperf pipeline (#3788)

- Referred to public documentation, source, and iterative attempts to create and improve build and test pipeline.
- ctest failures are due to the test node not having expected marketing name string and override not working.
- The fix should be on the omniperf repo side of things, so this pull request should be fine as is.

* External CI: create omniperf pipeline IDs, update nightly build (#3790)

* Fixed greater than to be less than in rocFFT changes

* fix footnote for 6.1.0 (#3791)

* fix footnote for 6.1.0

* fix empty columns in historical KFD title

* External CI: Publish wheel as artifact for rocPyDecode (#3796)

* fix build rocal for ROCm6.2.1

* Add ROCm6.2.1 manifest file

* External CI: fix hip-tests symlink creation (#3799)

* Docs: Add Ubuntu 24.04.1 (#3801)

* add ubuntu 24.04.1

* add 24.04.1 to bottom os section

* fix heading and template

* Update compatibility-matrix.rst for OpenMP version

* Update compatibility-matrix-historical-6.0.csv for OpenMP version

* rm ubuntu 24.04.1 from 6.2.0

* Update docs/compatibility/compatibility-matrix.rst

Co-authored-by: Young Hui - AMD <145490163+yhuiYH@users.noreply.github.com>

* rm duplicate ubuntu in historical

---------

Co-authored-by: Young Hui - AMD <145490163+yhuiYH@users.noreply.github.com>

* Docs: Add Ubuntu 24.04.1 (#3801)

* add ubuntu 24.04.1

* add 24.04.1 to bottom os section

* fix heading and template

* Update compatibility-matrix.rst for OpenMP version

* Update compatibility-matrix-historical-6.0.csv for OpenMP version

* rm ubuntu 24.04.1 from 6.2.0

* Update docs/compatibility/compatibility-matrix.rst

Co-authored-by: Young Hui - AMD <145490163+yhuiYH@users.noreply.github.com>

* rm duplicate ubuntu in historical

---------

Co-authored-by: Young Hui - AMD <145490163+yhuiYH@users.noreply.github.com>

* External CI: fixes for rocMLIR and nightly build (#3800)

* External CI: fix symlinks for rocMLIR and nightly build

* add pipeline IDs for hip-tests

* fix hip-test ID typo

* remove llvm-alt license (#3727)

* remove llvm-alt license

* fix linting error

* External CI: enable ROCR-Runtime tests (#3809)

* External CI: default branches for hip-tests, omniperf (#3811)

* External CI: torch and torchvision smoke tests (#3810)

* External CI: torch and torchvision smoke tests

- Fixed issues with package name and version for the vision wheel that prevented it from installing. A patch is used until my pull request in vision repo is merged.
- Referred to rocAutomation scripts to pick which test scripts to run out of the many in the torch and vision repo, and iteratively tested suggested scripts to see which ones completed in a timely manner.
- Leveraging pytest-azurepipelines module to automatically parse and graph results from these tests.

* External CI: omnitrace build pipeline (#3812)

* External CI: omnitrace build pipeline starter

- Adding initial set of dependencies and build flags.

* External CI: omnitrace build pipeline

- Add bison, rccl, texinfo dependencies based on build failures.
- Add AMDGPU_TARGETS flag
- Add ROCm binaries to PATH for clang-format and other tools used.

* Fix indentation

---------

Co-authored-by: Daniel Su <danielsu@amd.com>

* External CI: AMDMIGraphX Build Fix (#3814)

- Swap to default gcc on OS to resolve build errors from recent commits.
- Added libdnnl-dev dependency from iterative attempts with compiler change.
- Referred to the passing GitHub checks to observe the compilers that was used.
- Build CK jit lib and include in AMDMIGraphX build.

* External CI: test fixes w/ roctracer, list omniperf as partially succeeding (#3815)

* External CI: rpp tests (#3816)

* External CI: Build pipeline for rocprofiler-sdk (#3819)

* External CI: Pipeline for rocprofiler-sdk

* Add rocprofiler dependency

* External CI: rocprofiler-sdk build pipeline

---------

Co-authored-by: Daniel Su <danielsu@amd.com>

* External CI: Fix/add missing pipeline IDs (#3818)

* Update default.xml - Change 6.2.1 to 6.2.2

* Add ROCm6.2.1 manifest file

* External CI: omnitrace tests (#3822)

* Update tags to 6.2.2 (#3827)

* Update tags to 6.2.2 (#3827)

* External CI: add roctracer to roc/hipSOLVER test deps (#3825)

* External CI: add rocprofiler-sdk pipeline IDs (#3824)

* External CI: AMDMIGraphX Smoke Tests (#3830)

Co-authored-by: Daniel Su <danielsu@amd.com>

* External CI: MIOpen tests (#3837)

* Point to release history instead of deprecated changelog (#3836)

* External CI: filter out hipTensor extended tests (#3838)

* added revised note re. radeon gpus (#3839)

* Restructured the contributions section. (#3715)

* testing if this file is editable

* changed 'kebob-case' to 'dash-case'

* Restructured the page to be more straightforward and provide additional repo information

* forgot to save

* Moved the topic sentence

* Wrong accent on the a in diataxis

* Removed the feedback info from contributing and moved it to Feedback

* fixed spelling errors

* fixed some wording and removed second person text

* consolidated Build and Structure into Contribute; edited toolchai to (hopefully) conform to style guide; updated toc

* updated the titles in the toc

* made changes based on feedback

* it's better when you save

* removed structure and build; fixed something for the linter

* added rst to wordlist

* added customizations to wordlist

* Add links to gpu cluster network guides (#3763)

* Add links to gpu cluster network guides

* Add newline character to eof

* Make link absolute

* add dynamic branch in toc

* remove unnecessary page

clean up

* clean up index/toc

* make multi-node topics adjacent

---------

Co-authored-by: Peter Park <peter.park@amd.com>

* Point to release history instead of deprecated changelog (#3836)

* Restructured the contributions section. (#3715)

* testing if this file is editable

* changed 'kebob-case' to 'dash-case'

* Restructured the page to be more straightforward and provide additional repo information

* forgot to save

* Moved the topic sentence

* Wrong accent on the a in diataxis

* Removed the feedback info from contributing and moved it to Feedback

* fixed spelling errors

* fixed some wording and removed second person text

* consolidated Build and Structure into Contribute; edited toolchai to (hopefully) conform to style guide; updated toc

* updated the titles in the toc

* made changes based on feedback

* it's better when you save

* removed structure and build; fixed something for the linter

* added rst to wordlist

* added customizations to wordlist

* Add links to gpu cluster network guides (#3763)

* Add links to gpu cluster network guides

* Add newline character to eof

* Make link absolute

* add dynamic branch in toc

* remove unnecessary page

clean up

* clean up index/toc

* make multi-node topics adjacent

---------

Co-authored-by: Peter Park <peter.park@amd.com>

* updated the radeon note (#3850)

* External CI: Fix rocPyDecode wheel creation (#3852)

- Set values for expected environment variables.
- Accompanying changes required in rocPyDecode repo. Pull request will be made.

* External CI: pytorch vision patch removal (#3855)

My pull request applying this patch was merged upstream, so this is no longer needed and will break the pipeline since it can no longer be applied.

* Build(deps): Bump rocm-docs-core from 1.8.1 to 1.8.2 in /docs/sphinx (#3807)

Bumps [rocm-docs-core](https://github.com/ROCm/rocm-docs-core) from 1.8.1 to 1.8.2.
- [Release notes](https://github.com/ROCm/rocm-docs-core/releases)
- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/v1.8.2/CHANGELOG.md)
- [Commits](https://github.com/ROCm/rocm-docs-core/compare/v1.8.1...v1.8.2)

---
updated-dependencies:
- dependency-name: rocm-docs-core
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* updated the radeon note, as it were (#3857)

* updated the radeon note, as it were

* updated the note again

* Set devops team as codeowners for rocm-build (#3860)

* Set ext CI as codeowners for rocm-build

* Update CODEOWNERS to rocm-devops

* External CI: Add option to pull mainline branch for dependencies (#3689)

* External CI: Add option to pull mainline branch for dependencies

* Missing parameter for mainline branch dependencies.

* External CI: mainline branch definitions

* Removed MIGraphX optimization page (#3848)

* External CI: add a global variable to control gfx942 tests (#3864)

* External CI: update component default/mainline branches (#3871)

* External CI: Stop building gfx90a (#3872)

Save on VM resources until infrastructure has test targets.

* External CI: add libstdc++-12 to rocMLIR (#3874)

* Add building doc section (#3873)

* External CI: programmatically get latest aqlprofile (#3876)

* External CI: use ctest for rocm-examples (#3877)

* External CI: Tensile pipeline (#3884)

* add oversubscription conceptual doc (#3885)

add mitigiation steps

add to toc

move page for build

move doc

fix spelling

update doc

update oversubscription

update order

fix spelling

add oversubscription to wordlist

move oversubscription topic to bottom of toc and index

* add oversubscription conceptual doc (#3885)

add mitigiation steps

add to toc

move page for build

move doc

fix spelling

update doc

update oversubscription

update order

fix spelling

add oversubscription to wordlist

move oversubscription topic to bottom of toc and index

(cherry picked from commit d0ecf51b0c)

* add oversubscription conceptual doc (#3885)

(cherry picked from commit d0ecf51b0c)

* Add building doc section (#3873)

(cherry picked from commit abc0e6a087)

* External CI: Add pipeline to build upstream boost (#3896)

* Update bitsandbytes branch in docs (#3898)

* Update bitsandbytes branch in docs (#3898)

(cherry picked from commit b541be7bcb)

* Documentation: Add reference to precision-support floating-point types (#3899)

* External CI: use Boost template for MIOpen (#3903)

* External CI: create rocprofiler-systems pipeline (#3906)

* External CI: omnitrace/rocprof-sys pipeline IDs (#3908)

* External CI: MIOpen parse test results (#3913)

* External CI: Use pip to install latest cmake on test system (#3915)

* added a link to the compatibility matrix (#3904)

* added a link to the compatibility matrix

* removed quotes

* docs: Remove invalid amd_iommu=on parameter

Per kernel-parameters.txt, there is no "on" option for amd_iommu. While
intel_iommu has it, amd_iommu is automatically on unless specified
otherwise. For more info, see these 2 links:

https://www.kernel.org/doc/Documentation/admin-guide/kernel-parameters.txt
75aa74d52f/drivers/iommu/amd/init.c (L3481)

Signed-off-by: Kent Russell <kent.russell@amd.com>

* docs: Remove invalid amd_iommu=on parameter

Per kernel-parameters.txt, there is no "on" option for amd_iommu. While
intel_iommu has it, amd_iommu is automatically on unless specified
otherwise. For more info, see these 2 links:

https://www.kernel.org/doc/Documentation/admin-guide/kernel-parameters.txt
75aa74d52f/drivers/iommu/amd/init.c (L3481)

Signed-off-by: Kent Russell <kent.russell@amd.com>
(cherry picked from commit 74333b667d)

* External CI: hipBLASLt build now requires python packaging module (#3926)

https://github.com/ROCm/hipBLASLt/pull/1250/files#diff-fee2e6f068b33fca3a1dc49392de8848dbf05c3f4632b680abb1052523e5a30fR35

* External CI: Moved location of upstream pytorch build scripts (#3930)

https://github.com/pytorch/pytorch/pull/138103

* External CI: disable rocMLIR tests (#3931)

* External CI: disable rocMLIR tests

* roctracer AMDGPU_TARGETS flag

* External CI: create a GPU diagnostics template (#3932)

* External CI: Add CK into pytorch build environment (#3934)

* Update rocm-6.2.2.xml (#3927)

vim typo removed

* External CI: add support to disable individual component tests (#3938)

* External CI: AMDMIGraphX greater-equal pip dependencies (#3939)

* Build(deps): Bump rocm-docs-core from 1.8.2 to 1.8.3 in /docs/sphinx (#3933)

Bumps [rocm-docs-core](https://github.com/ROCm/rocm-docs-core) from 1.8.2 to 1.8.3.
- [Release notes](https://github.com/ROCm/rocm-docs-core/releases)
- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/develop/CHANGELOG.md)
- [Commits](https://github.com/ROCm/rocm-docs-core/compare/v1.8.2...v1.8.3)

---
updated-dependencies:
- dependency-name: rocm-docs-core
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* External CI: rocDecode add libva-amdgpu-dev dependency (#3940)

* External CI: enumerate GPUs in gpu-diagnostics (#3942)

* External CI: move gpu-diag directly before tests (#3943)

* External CI: fix HIP_PIPELINE_ID (#3944)

* External CI: pytorch pipeline updates (#3948)

To support recent upstream changes and issues observed.

* External CI: rocpydecode dependency installation change (#3954)

- Install pybind11 through pip instead of apt
- Add pip-installed pybind11 path to CMAKE_PREFIX_PATH
- Tested against source of PR 122

* External CI: do not assume python is python3 for rocpydecode (#3955)

* Improve consistency of the gpu-arch-specs table. (#3936)

* Improve consistency of the gpu-arch-specs table.

* Add XCD to the glossary.

* External CI: Always force rocPyDecode cleanup step

* External CI: Add aqlprofile to Tensile test dependencies (#3961)

* add vllm performance validation doc (#3964)

* External CI: various fixes (#3963)

* add suggestions to vllm perf validation doc (#3968)

* External CI: move allowPartiallySucceededBuilds to library variable (#3970)

* External CI: suppress GPU diag warnings (#3972)

* External CI: rocprofiler-compute pipeline files (#3973)

* External CI: disable reload AMDGPU (#3974)

* Update links to vllm perf validation doc (#3971)

* update links to vllm perf validation doc

* add PagedAttention to wordlist

* External CI: Change test setup for rocPyDecode (#3978)

- Use multiple potential locations for pybind11 to be found by cmake.

* External CI: add roctracer to rocBLAS deps (#3982)

* External CI: decode test changes (#3983)

- Only target container with access to first device
- Ensure pybind11-dev is uninstalled before the package manager install steps

* Changed the introductory text linked to Radeon (#3988)

Co-authored-by: prbasyal <prbasyal@amd.com>

* External CI: finish rocprofiler-compute enablement (#3995)

* External CI: add aomp as rocprofiler-systems dependency (#3996)

* External CI: remove omniperf from nightly (#4000)

* Sync from internal develop 6.2.4 (#4002)

* add radeon pro v710 to gpu arch specs (#192)

* Add V710 specs

gpg:                using RSA key
22223038B47B3ED4B3355AB11B54779B4780494E
gpg: Good signature from "Peter Park (MKMPETEPARK01)
<peter.park@amd.com>" [ultimate]
add some specs

add cols

clean up extra line

* fix graphics l1 cache description

* update SGPR for RDNA2 and RDNA3 archs

* update VGPR

* Apply suggestions from code review

* change l2 cache to 4

* Update docs/reference/gpu-arch-specs.rst

* ROCm 6.2.4 compatibility matrix (#186)

* prep compat column (historical) and mi300x column

* update historical compat matrix for 6.2.4

* update compat matrix for 6.2.4

* fix compat

* fix thunk version

* fix hipify ver

* ROCm 6.2.4 release notes (#184)

* prep 6.2.4 release notes

* add mathlibs

* add detail component changes

* rm non-updated linnks

* fix sentence

* fix rocthrust v

* rm offline installer

* condense

* add leo/ram fdback

words

* update documentation section

* add rocm on radeon note

* update os support note wording

* update release

* update version and GA date to 10-17

* update 6.2.4 rn

* update wording

* add link to v710

* update wording

* update templ

* simplify note

* words

os note

words

* change URLs to latest

* update link to supported GPUs

* Update versions.md 6.2.4 date to Oct 18

* Update conf.py release note date to Oct 18

---------

Co-authored-by: Sam Wu <22262939+samjwu@users.noreply.github.com>

* Sync change from ROCm to ROCm-internal (#194)

* Fix Radeon link and point at R6.1.3 as absolute link (#3757)

* Update ROCm manifest to 6.2.1

* Update ROCm branch name

* Add 6.2.1 to version list (#3770)

* Add links to GH issues in 6.2.1 release notes (#3769)

* add MAD page

* link to GitHub issues in release notes known issues

* update templates for 6.2.1

* Revert "add MAD page"

This reverts commit 9cce72bba3.

* update wordlist for spellcheck linter

* add rccl note

* update rocal version change heading to be more obvious

* make rocal note more specific

* fix missing space

* fix capitalization

* Update RCCL known issue wording (#3775)

* add MAD page

* fix wording in RCCL known issue

* Revert "add MAD page"

This reverts commit c81d0f3b0a.

* update llvm version for 6.2.1 (#3779)

* Fix broken links in 6.2.1 release notes (#3782)

* External CI: Replace libomp dependencies with aomp (#3781)

Add roctracer dependency for hipBLAS and rocWMMA testing

* External CI: Add rocprofiler v1 and v2 smoke tests (#3784)

* External CI: ROCgdb smoke tests (#3785)

- Since this is an autotools project and not cmake, build and test on gfx942 system instead of separating into two jobs. Pipeline time is short anyway.
- Follow build instructions to update build flags and to incorporate the ROCdbgapi.
- Results are not parsed and graphed, but the log contents are printed at the end. This was helpful for debugging and will be kept in the pipeline, as the make check-gdb command's output was not helpful on its own.

* External CI: rocPyDecode Smoke Test (#3786)

* External CI: omniperf pipeline (#3788)

- Referred to public documentation, source, and iterative attempts to create and improve build and test pipeline.
- ctest failures are due to the test node not having expected marketing name string and override not working.
- The fix should be on the omniperf repo side of things, so this pull request should be fine as is.

* External CI: create omniperf pipeline IDs, update nightly build (#3790)

* Fixed greater than to be less than in rocFFT changes

* fix footnote for 6.1.0 (#3791)

* fix footnote for 6.1.0

* fix empty columns in historical KFD title

* External CI: Publish wheel as artifact for rocPyDecode (#3796)

* External CI: fix hip-tests symlink creation (#3799)

* Docs: Add Ubuntu 24.04.1 (#3801)

* add ubuntu 24.04.1

* add 24.04.1 to bottom os section

* fix heading and template

* Update compatibility-matrix.rst for OpenMP version

* Update compatibility-matrix-historical-6.0.csv for OpenMP version

* rm ubuntu 24.04.1 from 6.2.0

* Update docs/compatibility/compatibility-matrix.rst

Co-authored-by: Young Hui - AMD <145490163+yhuiYH@users.noreply.github.com>

* rm duplicate ubuntu in historical

---------

Co-authored-by: Young Hui - AMD <145490163+yhuiYH@users.noreply.github.com>

* External CI: fixes for rocMLIR and nightly build (#3800)

* External CI: fix symlinks for rocMLIR and nightly build

* add pipeline IDs for hip-tests

* fix hip-test ID typo

* remove llvm-alt license (#3727)

* remove llvm-alt license

* fix linting error

* External CI: enable ROCR-Runtime tests (#3809)

* External CI: default branches for hip-tests, omniperf (#3811)

* External CI: torch and torchvision smoke tests (#3810)

* External CI: torch and torchvision smoke tests

- Fixed issues with package name and version for the vision wheel that prevented it from installing. A patch is used until my pull request in vision repo is merged.
- Referred to rocAutomation scripts to pick which test scripts to run out of the many in the torch and vision repo, and iteratively tested suggested scripts to see which ones completed in a timely manner.
- Leveraging pytest-azurepipelines module to automatically parse and graph results from these tests.

* External CI: omnitrace build pipeline (#3812)

* External CI: omnitrace build pipeline starter

- Adding initial set of dependencies and build flags.

* External CI: omnitrace build pipeline

- Add bison, rccl, texinfo dependencies based on build failures.
- Add AMDGPU_TARGETS flag
- Add ROCm binaries to PATH for clang-format and other tools used.

* Fix indentation

---------

Co-authored-by: Daniel Su <danielsu@amd.com>

* External CI: AMDMIGraphX Build Fix (#3814)

- Swap to default gcc on OS to resolve build errors from recent commits.
- Added libdnnl-dev dependency from iterative attempts with compiler change.
- Referred to the passing GitHub checks to observe the compilers that was used.
- Build CK jit lib and include in AMDMIGraphX build.

* External CI: test fixes w/ roctracer, list omniperf as partially succeeding (#3815)

* External CI: rpp tests (#3816)

* External CI: Build pipeline for rocprofiler-sdk (#3819)

* External CI: Pipeline for rocprofiler-sdk

* Add rocprofiler dependency

* External CI: rocprofiler-sdk build pipeline

---------

Co-authored-by: Daniel Su <danielsu@amd.com>

* External CI: Fix/add missing pipeline IDs (#3818)

* External CI: omnitrace tests (#3822)

* Update tags to 6.2.2 (#3827)

* External CI: add roctracer to roc/hipSOLVER test deps (#3825)

* External CI: add rocprofiler-sdk pipeline IDs (#3824)

* External CI: AMDMIGraphX Smoke Tests (#3830)

Co-authored-by: Daniel Su <danielsu@amd.com>

* External CI: MIOpen tests (#3837)

* Point to release history instead of deprecated changelog (#3836)

* External CI: filter out hipTensor extended tests (#3838)

* added revised note re. radeon gpus (#3839)

* Restructured the contributions section. (#3715)

* testing if this file is editable

* changed 'kebob-case' to 'dash-case'

* Restructured the page to be more straightforward and provide additional repo information

* forgot to save

* Moved the topic sentence

* Wrong accent on the a in diataxis

* Removed the feedback info from contributing and moved it to Feedback

* fixed spelling errors

* fixed some wording and removed second person text

* consolidated Build and Structure into Contribute; edited toolchai to (hopefully) conform to style guide; updated toc

* updated the titles in the toc

* made changes based on feedback

* it's better when you save

* removed structure and build; fixed something for the linter

* added rst to wordlist

* added customizations to wordlist

* Add links to gpu cluster network guides (#3763)

* Add links to gpu cluster network guides

* Add newline character to eof

* Make link absolute

* add dynamic branch in toc

* remove unnecessary page

clean up

* clean up index/toc

* make multi-node topics adjacent

---------

Co-authored-by: Peter Park <peter.park@amd.com>

* updated the radeon note (#3850)

* External CI: Fix rocPyDecode wheel creation (#3852)

- Set values for expected environment variables.
- Accompanying changes required in rocPyDecode repo. Pull request will be made.

* External CI: pytorch vision patch removal (#3855)

My pull request applying this patch was merged upstream, so this is no longer needed and will break the pipeline since it can no longer be applied.

* Build(deps): Bump rocm-docs-core from 1.8.1 to 1.8.2 in /docs/sphinx (#3807)

Bumps [rocm-docs-core](https://github.com/ROCm/rocm-docs-core) from 1.8.1 to 1.8.2.
- [Release notes](https://github.com/ROCm/rocm-docs-core/releases)
- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/v1.8.2/CHANGELOG.md)
- [Commits](https://github.com/ROCm/rocm-docs-core/compare/v1.8.1...v1.8.2)

---
updated-dependencies:
- dependency-name: rocm-docs-core
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* updated the radeon note, as it were (#3857)

* updated the radeon note, as it were

* updated the note again

* Set devops team as codeowners for rocm-build (#3860)

* Set ext CI as codeowners for rocm-build

* Update CODEOWNERS to rocm-devops

* External CI: Add option to pull mainline branch for dependencies (#3689)

* External CI: Add option to pull mainline branch for dependencies

* Missing parameter for mainline branch dependencies.

* External CI: mainline branch definitions

* Removed MIGraphX optimization page (#3848)

* External CI: add a global variable to control gfx942 tests (#3864)

* External CI: update component default/mainline branches (#3871)

* External CI: Stop building gfx90a (#3872)

Save on VM resources until infrastructure has test targets.

* External CI: add libstdc++-12 to rocMLIR (#3874)

* Add building doc section (#3873)

* External CI: programmatically get latest aqlprofile (#3876)

* External CI: use ctest for rocm-examples (#3877)

* External CI: Tensile pipeline (#3884)

* add oversubscription conceptual doc (#3885)

add mitigiation steps

add to toc

move page for build

move doc

fix spelling

update doc

update oversubscription

update order

fix spelling

add oversubscription to wordlist

move oversubscription topic to bottom of toc and index

* add oversubscription conceptual doc (#3885)

(cherry picked from commit d0ecf51b0c)

* External CI: Add pipeline to build upstream boost (#3896)

* Update bitsandbytes branch in docs (#3898)

* Documentation: Add reference to precision-support floating-point types (#3899)

* External CI: use Boost template for MIOpen (#3903)

* External CI: create rocprofiler-systems pipeline (#3906)

* External CI: omnitrace/rocprof-sys pipeline IDs (#3908)

* External CI: MIOpen parse test results (#3913)

* External CI: Use pip to install latest cmake on test system (#3915)

* added a link to the compatibility matrix (#3904)

* added a link to the compatibility matrix

* removed quotes

* docs: Remove invalid amd_iommu=on parameter

Per kernel-parameters.txt, there is no "on" option for amd_iommu. While
intel_iommu has it, amd_iommu is automatically on unless specified
otherwise. For more info, see these 2 links:

https://www.kernel.org/doc/Documentation/admin-guide/kernel-parameters.txt
75aa74d52f/drivers/iommu/amd/init.c (L3481)

Signed-off-by: Kent Russell <kent.russell@amd.com>

* External CI: hipBLASLt build now requires python packaging module (#3926)

https://github.com/ROCm/hipBLASLt/pull/1250/files#diff-fee2e6f068b33fca3a1dc49392de8848dbf05c3f4632b680abb1052523e5a30fR35

* External CI: Moved location of upstream pytorch build scripts (#3930)

https://github.com/pytorch/pytorch/pull/138103

* External CI: disable rocMLIR tests (#3931)

* External CI: disable rocMLIR tests

* roctracer AMDGPU_TARGETS flag

* External CI: create a GPU diagnostics template (#3932)

* External CI: Add CK into pytorch build environment (#3934)

* External CI: add support to disable individual component tests (#3938)

* External CI: AMDMIGraphX greater-equal pip dependencies (#3939)

* Build(deps): Bump rocm-docs-core from 1.8.2 to 1.8.3 in /docs/sphinx (#3933)

Bumps [rocm-docs-core](https://github.com/ROCm/rocm-docs-core) from 1.8.2 to 1.8.3.
- [Release notes](https://github.com/ROCm/rocm-docs-core/releases)
- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/develop/CHANGELOG.md)
- [Commits](https://github.com/ROCm/rocm-docs-core/compare/v1.8.2...v1.8.3)

---
updated-dependencies:
- dependency-name: rocm-docs-core
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* External CI: rocDecode add libva-amdgpu-dev dependency (#3940)

* External CI: enumerate GPUs in gpu-diagnostics (#3942)

* External CI: move gpu-diag directly before tests (#3943)

* External CI: fix HIP_PIPELINE_ID (#3944)

---------

Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Kent Russell <kent.russell@amd.com>
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>
Co-authored-by: Sam Wu <22262939+samjwu@users.noreply.github.com>
Co-authored-by: Wang, Yanyao <yanyao.wang@amd.com>
Co-authored-by: Yanyao Wang <yanywang@amd.com>
Co-authored-by: Peter Park <peter.park@amd.com>
Co-authored-by: Young Hui - AMD <145490163+yhuiYH@users.noreply.github.com>
Co-authored-by: Joseph Macaranas <145489236+amd-jmacaran@users.noreply.github.com>
Co-authored-by: Daniel Su <danielsu@amd.com>
Co-authored-by: Sandra Polifroni <sandra.polifroni@amd.com>
Co-authored-by: randyh62 <42045079+randyh62@users.noreply.github.com>
Co-authored-by: Michael Benavidez <michael.benavidez@amd.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: MKKnorr <MKKnorr@web.de>
Co-authored-by: Kent Russell <kent.russell@amd.com>
Co-authored-by: Joseph Greathouse <jlgreathouse@users.noreply.github.com>

* 6.2.4 release notes: add known/fixed issues (#193)

* add "for compute workloads" wording for clarity

* add AMDSMI resolved issue

* add dlm known issue

intro text

wording

* update wording

rm bullet point

update wording

* fix spellcheck due to spacing

* rm s

* rm gfx1151

* remove dlm known issue

* update list of updated docs; note for Radeon users

fmt

* update GA date for 6.2.4

* fix rdc version

* fix RDC version strings (#196)

* revert outdataed change for .azuredevops

* Fix 6.2.4 date in versions.md

Co-authored-by: Sam Wu <22262939+samjwu@users.noreply.github.com>

---------

Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Kent Russell <kent.russell@amd.com>
Co-authored-by: Peter Park <peter.park@amd.com>
Co-authored-by: Sam Wu <22262939+samjwu@users.noreply.github.com>
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>
Co-authored-by: Wang, Yanyao <yanyao.wang@amd.com>
Co-authored-by: Yanyao Wang <yanywang@amd.com>
Co-authored-by: Young Hui - AMD <145490163+yhuiYH@users.noreply.github.com>
Co-authored-by: Joseph Macaranas <145489236+amd-jmacaran@users.noreply.github.com>
Co-authored-by: Daniel Su <danielsu@amd.com>
Co-authored-by: Sandra Polifroni <sandra.polifroni@amd.com>
Co-authored-by: randyh62 <42045079+randyh62@users.noreply.github.com>
Co-authored-by: Michael Benavidez <michael.benavidez@amd.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: MKKnorr <MKKnorr@web.de>
Co-authored-by: Kent Russell <kent.russell@amd.com>
Co-authored-by: Joseph Greathouse <jlgreathouse@users.noreply.github.com>

* fix links in release notes 6.2.4 (#4008)

* Remove extra line

* Update xml files for 6.2.4 (#4012)

* Update xml files for 6.2.4

* Update README with 6.2.4

* Increase visibility of programming guide

* Docs: Update what is rocm description

* Apply suggestions from code review

Co-authored-by: randyh62 <42045079+randyh62@users.noreply.github.com>

* Update docs/how-to/hip_programming_guide.rst

Co-authored-by: MKKnorr <MKKnorr@web.de>

* WIP

* Update docs/index.md

* Update docs/how-to/hip_programming_guide.rst

Co-authored-by: MKKnorr <MKKnorr@web.de>

* Update docs/how-to/programming_guide.rst

* Update docs/what-is-rocm.rst

* Apply suggestions from code review

Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com>

* Update docs/how-to/programming_guide.rst

Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com>

* Remove tip

* External CI: allow test failures to present as failures on Github (#3993)

* External CI: disable rdmatest and rocrtstFunc.Memory_Max_Mem (#4016)

* Added 6.2.4 manifest.xml

* External CI: fix comgr build (#4025)

* External CI: increase Tensile test timeout to 90 mins (#4027)

---------

Signed-off-by: David Galiffi <David.Galiffi@amd.com>
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Kent Russell <kent.russell@amd.com>
Co-authored-by: Sam Wu <22262939+samjwu@users.noreply.github.com>
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>
Co-authored-by: Peter Park <peter.park@amd.com>
Co-authored-by: Yanyao Wang <yanywang@amd.com>
Co-authored-by: Wang, Yanyao <yanyao.wang@amd.com>
Co-authored-by: David Galiffi <dgaliffi@amd.com>
Co-authored-by: Chris Kime <Christopher.Kime@amd.com>
Co-authored-by: ozziemoreno <109979778+ozziemoreno@users.noreply.github.com>
Co-authored-by: Sandra Polifroni <sandra.polifroni@amd.com>
Co-authored-by: Young Hui - AMD <145490163+yhuiYH@users.noreply.github.com>
Co-authored-by: Joseph Macaranas <145489236+amd-jmacaran@users.noreply.github.com>
Co-authored-by: Daniel Su <danielsu@amd.com>
Co-authored-by: randyh62 <42045079+randyh62@users.noreply.github.com>
Co-authored-by: JeniferC99 <150404595+JeniferC99@users.noreply.github.com>
Co-authored-by: Michael Benavidez <michael.benavidez@amd.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: MKKnorr <MKKnorr@web.de>
Co-authored-by: Kent Russell <kent.russell@amd.com>
Co-authored-by: Joseph Greathouse <jlgreathouse@users.noreply.github.com>
Co-authored-by: Johannes Maria Frank <jmfrank63@gmail.com>
Co-authored-by: Brian Cornille <bcornill@amd.com>
Co-authored-by: Joseph Macaranas <Joseph.Macaranas@amd.com>
Co-authored-by: Pratik Basyal <pratik.basyal@amd.com>
Co-authored-by: prbasyal <prbasyal@amd.com>
Co-authored-by: Istvan Kiss <neon60@gmail.com>
Co-authored-by: Leo Paoletti <164940351+lpaoletti@users.noreply.github.com>
Co-authored-by: Ameya Keshava Mallya <ameyakeshava.mallya@amd.com>
2024-11-14 13:14:37 -05:00
srawat
c1f0a8d9d9 Adding TransferBench into list of components (#185) 2024-11-08 15:46:59 -07:00
Sam Wu
c9754cb9d8 Test Sphinx Sitemap (#162)
* Use sphinx-sitemap

* Update baseurl for sphinx-sitemap

* Regenerate doc reqs
2024-11-08 15:46:04 -07:00
Peter Park
b7ecf6d552 Rename Omnitools to ROCm Compute/Systems Profiler (#183)
* rename Omniperf and Omnitrace

* rename labels

rename more labels

* update licenses and rocm-tools.md

* fix rocprof-sys ref
2024-11-07 18:01:26 -05:00
Jeffrey Novotny
48d2d16563 Remove deprecated architectures from LLVM targets in GPU architecture specs (#198) 2024-11-07 12:33:28 -05:00
Peter Park
47cda1bc70 fix RDC version strings (#196) 2024-11-06 13:51:36 -05:00
Peter Park
34c0266496 6.2.4 release notes: add known/fixed issues (#193)
* add "for compute workloads" wording for clarity

* add AMDSMI resolved issue

* add dlm known issue

intro text

wording

* update wording

rm bullet point

update wording

* fix spellcheck due to spacing

* rm s

* rm gfx1151

* remove dlm known issue

* update list of updated docs; note for Radeon users

fmt

* update GA date for 6.2.4

* fix rdc version
2024-11-06 12:42:55 -05:00
alexxu-amd
8e3d51c31d Sync change from ROCm to ROCm-internal (#194)
* Fix Radeon link and point at R6.1.3 as absolute link (#3757)

* Update ROCm manifest to 6.2.1

* Update ROCm branch name

* Add 6.2.1 to version list (#3770)

* Add links to GH issues in 6.2.1 release notes (#3769)

* add MAD page

* link to GitHub issues in release notes known issues

* update templates for 6.2.1

* Revert "add MAD page"

This reverts commit 9cce72bba3.

* update wordlist for spellcheck linter

* add rccl note

* update rocal version change heading to be more obvious

* make rocal note more specific

* fix missing space

* fix capitalization

* Update RCCL known issue wording (#3775)

* add MAD page

* fix wording in RCCL known issue

* Revert "add MAD page"

This reverts commit c81d0f3b0a.

* update llvm version for 6.2.1 (#3779)

* Fix broken links in 6.2.1 release notes (#3782)

* External CI: Replace libomp dependencies with aomp (#3781)

Add roctracer dependency for hipBLAS and rocWMMA testing

* External CI: Add rocprofiler v1 and v2 smoke tests (#3784)

* External CI: ROCgdb smoke tests (#3785)

- Since this is an autotools project and not cmake, build and test on gfx942 system instead of separating into two jobs. Pipeline time is short anyway.
- Follow build instructions to update build flags and to incorporate the ROCdbgapi.
- Results are not parsed and graphed, but the log contents are printed at the end. This was helpful for debugging and will be kept in the pipeline, as the make check-gdb command's output was not helpful on its own.

* External CI: rocPyDecode Smoke Test (#3786)

* External CI: omniperf pipeline (#3788)

- Referred to public documentation, source, and iterative attempts to create and improve build and test pipeline.
- ctest failures are due to the test node not having expected marketing name string and override not working.
- The fix should be on the omniperf repo side of things, so this pull request should be fine as is.

* External CI: create omniperf pipeline IDs, update nightly build (#3790)

* Fixed greater than to be less than in rocFFT changes

* fix footnote for 6.1.0 (#3791)

* fix footnote for 6.1.0

* fix empty columns in historical KFD title

* External CI: Publish wheel as artifact for rocPyDecode (#3796)

* External CI: fix hip-tests symlink creation (#3799)

* Docs: Add Ubuntu 24.04.1 (#3801)

* add ubuntu 24.04.1

* add 24.04.1 to bottom os section

* fix heading and template

* Update compatibility-matrix.rst for OpenMP version

* Update compatibility-matrix-historical-6.0.csv for OpenMP version

* rm ubuntu 24.04.1 from 6.2.0

* Update docs/compatibility/compatibility-matrix.rst

Co-authored-by: Young Hui - AMD <145490163+yhuiYH@users.noreply.github.com>

* rm duplicate ubuntu in historical

---------

Co-authored-by: Young Hui - AMD <145490163+yhuiYH@users.noreply.github.com>

* External CI: fixes for rocMLIR and nightly build (#3800)

* External CI: fix symlinks for rocMLIR and nightly build

* add pipeline IDs for hip-tests

* fix hip-test ID typo

* remove llvm-alt license (#3727)

* remove llvm-alt license

* fix linting error

* External CI: enable ROCR-Runtime tests (#3809)

* External CI: default branches for hip-tests, omniperf (#3811)

* External CI: torch and torchvision smoke tests (#3810)

* External CI: torch and torchvision smoke tests

- Fixed issues with package name and version for the vision wheel that prevented it from installing. A patch is used until my pull request in vision repo is merged.
- Referred to rocAutomation scripts to pick which test scripts to run out of the many in the torch and vision repo, and iteratively tested suggested scripts to see which ones completed in a timely manner.
- Leveraging pytest-azurepipelines module to automatically parse and graph results from these tests.

* External CI: omnitrace build pipeline (#3812)

* External CI: omnitrace build pipeline starter

- Adding initial set of dependencies and build flags.

* External CI: omnitrace build pipeline

- Add bison, rccl, texinfo dependencies based on build failures.
- Add AMDGPU_TARGETS flag
- Add ROCm binaries to PATH for clang-format and other tools used.

* Fix indentation

---------

Co-authored-by: Daniel Su <danielsu@amd.com>

* External CI: AMDMIGraphX Build Fix (#3814)

- Swap to default gcc on OS to resolve build errors from recent commits.
- Added libdnnl-dev dependency from iterative attempts with compiler change.
- Referred to the passing GitHub checks to observe the compilers that was used.
- Build CK jit lib and include in AMDMIGraphX build.

* External CI: test fixes w/ roctracer, list omniperf as partially succeeding (#3815)

* External CI: rpp tests (#3816)

* External CI: Build pipeline for rocprofiler-sdk (#3819)

* External CI: Pipeline for rocprofiler-sdk

* Add rocprofiler dependency

* External CI: rocprofiler-sdk build pipeline

---------

Co-authored-by: Daniel Su <danielsu@amd.com>

* External CI: Fix/add missing pipeline IDs (#3818)

* External CI: omnitrace tests (#3822)

* Update tags to 6.2.2 (#3827)

* External CI: add roctracer to roc/hipSOLVER test deps (#3825)

* External CI: add rocprofiler-sdk pipeline IDs (#3824)

* External CI: AMDMIGraphX Smoke Tests (#3830)

Co-authored-by: Daniel Su <danielsu@amd.com>

* External CI: MIOpen tests (#3837)

* Point to release history instead of deprecated changelog (#3836)

* External CI: filter out hipTensor extended tests (#3838)

* added revised note re. radeon gpus (#3839)

* Restructured the contributions section. (#3715)

* testing if this file is editable

* changed 'kebob-case' to 'dash-case'

* Restructured the page to be more straightforward and provide additional repo information

* forgot to save

* Moved the topic sentence

* Wrong accent on the a in diataxis

* Removed the feedback info from contributing and moved it to Feedback

* fixed spelling errors

* fixed some wording and removed second person text

* consolidated Build and Structure into Contribute; edited toolchai to (hopefully) conform to style guide; updated toc

* updated the titles in the toc

* made changes based on feedback

* it's better when you save

* removed structure and build; fixed something for the linter

* added rst to wordlist

* added customizations to wordlist

* Add links to gpu cluster network guides (#3763)

* Add links to gpu cluster network guides

* Add newline character to eof

* Make link absolute

* add dynamic branch in toc

* remove unnecessary page

clean up

* clean up index/toc

* make multi-node topics adjacent

---------

Co-authored-by: Peter Park <peter.park@amd.com>

* updated the radeon note (#3850)

* External CI: Fix rocPyDecode wheel creation (#3852)

- Set values for expected environment variables.
- Accompanying changes required in rocPyDecode repo. Pull request will be made.

* External CI: pytorch vision patch removal (#3855)

My pull request applying this patch was merged upstream, so this is no longer needed and will break the pipeline since it can no longer be applied.

* Build(deps): Bump rocm-docs-core from 1.8.1 to 1.8.2 in /docs/sphinx (#3807)

Bumps [rocm-docs-core](https://github.com/ROCm/rocm-docs-core) from 1.8.1 to 1.8.2.
- [Release notes](https://github.com/ROCm/rocm-docs-core/releases)
- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/v1.8.2/CHANGELOG.md)
- [Commits](https://github.com/ROCm/rocm-docs-core/compare/v1.8.1...v1.8.2)

---
updated-dependencies:
- dependency-name: rocm-docs-core
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* updated the radeon note, as it were (#3857)

* updated the radeon note, as it were

* updated the note again

* Set devops team as codeowners for rocm-build (#3860)

* Set ext CI as codeowners for rocm-build

* Update CODEOWNERS to rocm-devops

* External CI: Add option to pull mainline branch for dependencies (#3689)

* External CI: Add option to pull mainline branch for dependencies

* Missing parameter for mainline branch dependencies.

* External CI: mainline branch definitions

* Removed MIGraphX optimization page (#3848)

* External CI: add a global variable to control gfx942 tests (#3864)

* External CI: update component default/mainline branches (#3871)

* External CI: Stop building gfx90a (#3872)

Save on VM resources until infrastructure has test targets.

* External CI: add libstdc++-12 to rocMLIR (#3874)

* Add building doc section (#3873)

* External CI: programmatically get latest aqlprofile (#3876)

* External CI: use ctest for rocm-examples (#3877)

* External CI: Tensile pipeline (#3884)

* add oversubscription conceptual doc (#3885)

add mitigiation steps

add to toc

move page for build

move doc

fix spelling

update doc

update oversubscription

update order

fix spelling

add oversubscription to wordlist

move oversubscription topic to bottom of toc and index

* add oversubscription conceptual doc (#3885)

(cherry picked from commit d0ecf51b0c)

* External CI: Add pipeline to build upstream boost (#3896)

* Update bitsandbytes branch in docs (#3898)

* Documentation: Add reference to precision-support floating-point types (#3899)

* External CI: use Boost template for MIOpen (#3903)

* External CI: create rocprofiler-systems pipeline (#3906)

* External CI: omnitrace/rocprof-sys pipeline IDs (#3908)

* External CI: MIOpen parse test results (#3913)

* External CI: Use pip to install latest cmake on test system (#3915)

* added a link to the compatibility matrix (#3904)

* added a link to the compatibility matrix

* removed quotes

* docs: Remove invalid amd_iommu=on parameter

Per kernel-parameters.txt, there is no "on" option for amd_iommu. While
intel_iommu has it, amd_iommu is automatically on unless specified
otherwise. For more info, see these 2 links:

https://www.kernel.org/doc/Documentation/admin-guide/kernel-parameters.txt
75aa74d52f/drivers/iommu/amd/init.c (L3481)

Signed-off-by: Kent Russell <kent.russell@amd.com>

* External CI: hipBLASLt build now requires python packaging module (#3926)

https://github.com/ROCm/hipBLASLt/pull/1250/files#diff-fee2e6f068b33fca3a1dc49392de8848dbf05c3f4632b680abb1052523e5a30fR35

* External CI: Moved location of upstream pytorch build scripts (#3930)

https://github.com/pytorch/pytorch/pull/138103

* External CI: disable rocMLIR tests (#3931)

* External CI: disable rocMLIR tests

* roctracer AMDGPU_TARGETS flag

* External CI: create a GPU diagnostics template (#3932)

* External CI: Add CK into pytorch build environment (#3934)

* External CI: add support to disable individual component tests (#3938)

* External CI: AMDMIGraphX greater-equal pip dependencies (#3939)

* Build(deps): Bump rocm-docs-core from 1.8.2 to 1.8.3 in /docs/sphinx (#3933)

Bumps [rocm-docs-core](https://github.com/ROCm/rocm-docs-core) from 1.8.2 to 1.8.3.
- [Release notes](https://github.com/ROCm/rocm-docs-core/releases)
- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/develop/CHANGELOG.md)
- [Commits](https://github.com/ROCm/rocm-docs-core/compare/v1.8.2...v1.8.3)

---
updated-dependencies:
- dependency-name: rocm-docs-core
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* External CI: rocDecode add libva-amdgpu-dev dependency (#3940)

* External CI: enumerate GPUs in gpu-diagnostics (#3942)

* External CI: move gpu-diag directly before tests (#3943)

* External CI: fix HIP_PIPELINE_ID (#3944)

---------

Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Kent Russell <kent.russell@amd.com>
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>
Co-authored-by: Sam Wu <22262939+samjwu@users.noreply.github.com>
Co-authored-by: Wang, Yanyao <yanyao.wang@amd.com>
Co-authored-by: Yanyao Wang <yanywang@amd.com>
Co-authored-by: Peter Park <peter.park@amd.com>
Co-authored-by: Young Hui - AMD <145490163+yhuiYH@users.noreply.github.com>
Co-authored-by: Joseph Macaranas <145489236+amd-jmacaran@users.noreply.github.com>
Co-authored-by: Daniel Su <danielsu@amd.com>
Co-authored-by: Sandra Polifroni <sandra.polifroni@amd.com>
Co-authored-by: randyh62 <42045079+randyh62@users.noreply.github.com>
Co-authored-by: Michael Benavidez <michael.benavidez@amd.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: MKKnorr <MKKnorr@web.de>
Co-authored-by: Kent Russell <kent.russell@amd.com>
Co-authored-by: Joseph Greathouse <jlgreathouse@users.noreply.github.com>
2024-10-25 14:41:40 -04:00
Peter Park
b9ec9507db ROCm 6.2.4 release notes (#184)
* prep 6.2.4 release notes

* add mathlibs

* add detail component changes

* rm non-updated linnks

* fix sentence

* fix rocthrust v

* rm offline installer

* condense

* add leo/ram fdback

words

* update documentation section

* add rocm on radeon note

* update os support note wording

* update release

* update version and GA date to 10-17

* update 6.2.4 rn

* update wording

* add link to v710

* update wording

* update templ

* simplify note

* words

os note

words

* change URLs to latest

* update link to supported GPUs

* Update versions.md 6.2.4 date to Oct 18

* Update conf.py release note date to Oct 18

---------

Co-authored-by: Sam Wu <22262939+samjwu@users.noreply.github.com>
2024-10-18 12:21:38 -04:00
Peter Park
be19ac1071 ROCm 6.2.4 compatibility matrix (#186)
* prep compat column (historical) and mi300x column

* update historical compat matrix for 6.2.4

* update compat matrix for 6.2.4

* fix compat

* fix thunk version

* fix hipify ver
2024-10-18 11:39:09 -04:00
Peter Park
916b5513cf add radeon pro v710 to gpu arch specs (#192)
* Add V710 specs

gpg:                using RSA key
22223038B47B3ED4B3355AB11B54779B4780494E
gpg: Good signature from "Peter Park (MKMPETEPARK01)
<peter.park@amd.com>" [ultimate]
add some specs

add cols

clean up extra line

* fix graphics l1 cache description

* update SGPR for RDNA2 and RDNA3 archs

* update VGPR

* Apply suggestions from code review

* change l2 cache to 4

* Update docs/reference/gpu-arch-specs.rst
2024-10-18 11:33:36 -04:00
133 changed files with 3799 additions and 4126 deletions

View File

@@ -197,3 +197,4 @@ jobs:
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/test.yml
parameters:
componentName: MIOpen
testParameters: '-VV --output-on-failure --force-new-ctest-process --output-junit test_output.xml --exclude-regex test_rnn_seq_api'

View File

@@ -126,6 +126,7 @@ jobs:
componentName: comgr
extraBuildFlags: >-
-DCMAKE_PREFIX_PATH="$(Build.SourcesDirectory)/llvm/build;$(Build.SourcesDirectory)/amd/device-libs/build"
-DCOMGR_DISABLE_SPIRV=1
-DCMAKE_BUILD_TYPE=Release
cmakeBuildDir: 'amd/comgr/build'
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/test.yml

View File

@@ -0,0 +1,166 @@
parameters:
- name: checkoutRepo
type: string
default: 'self'
- name: checkoutRef
type: string
default: ''
- name: aptPackages
type: object
default:
- cmake
- python3-pip
- name: pipModules
type: object
default:
- astunparse==1.6.2
- colorlover
- dash>=1.12.0
- matplotlib
- numpy>=1.17.5
- pandas>=1.4.3
- pymongo
- pyyaml
- tabulate
- tqdm
- dash-svg
- dash-bootstrap-components
- kaleido
- setuptools
- plotille
- mock
- pytest
- pytest-cov
- pytest-xdist
- name: rocmDependencies
type: object
default:
- clr
- llvm-project
- rocm-cmake
- rocm-core
- rocminfo
- ROCR-Runtime
- rocprofiler
- rocprofiler-register
- roctracer
jobs:
- job: omniperf
variables:
- group: common
- template: /.azuredevops/variables-global.yml
pool:
vmImage: ${{ variables.BASE_BUILD_POOL }}
workspace:
clean: all
strategy:
matrix:
gfx942:
JOB_GPU_TARGET: gfx942
steps:
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-other.yml
parameters:
aptPackages: ${{ parameters.aptPackages }}
pipModules: ${{ parameters.pipModules }}
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
parameters:
checkoutRepo: ${{ parameters.checkoutRepo }}
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-aqlprofile.yml
parameters:
${{ if eq(parameters.checkoutRef, '') }}:
dependencySource: staging
${{ elseif ne(parameters.checkoutRef, '') }}:
dependencySource: tag-builds
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
parameters:
dependencyList: ${{ parameters.rocmDependencies }}
gpuTarget: $(JOB_GPU_TARGET)
# CI case: download latest default branch build
${{ if eq(parameters.checkoutRef, '') }}:
dependencySource: staging
# manual build case: triggered by ROCm/ROCm repo
${{ elseif ne(parameters.checkoutRef, '') }}:
dependencySource: tag-builds
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/build-cmake.yml
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-upload.yml
parameters:
gpuTarget: $(JOB_GPU_TARGET)
- job: omniperf_testing
dependsOn: omniperf
condition: and(succeeded(), eq(variables.ENABLE_GFX942_TESTS, 'true'), not(containsValue(split(variables.DISABLED_GFX942_TESTS, ','), variables['Build.DefinitionName'])))
variables:
- group: common
- template: /.azuredevops/variables-global.yml
- name: PYTHON_VERSION
value: 3.10
pool: $(JOB_TEST_POOL)
workspace:
clean: all
strategy:
matrix:
gfx942:
JOB_GPU_TARGET: gfx942
JOB_TEST_POOL: ${{ variables.GFX942_TEST_POOL }}
steps:
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-other.yml
parameters:
aptPackages: ${{ parameters.aptPackages }}
pipModules: ${{ parameters.pipModules }}
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/preamble.yml
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
parameters:
checkoutRepo: ${{ parameters.checkoutRepo }}
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/local-artifact-download.yml
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-aqlprofile.yml
parameters:
${{ if eq(parameters.checkoutRef, '') }}:
dependencySource: staging
${{ elseif ne(parameters.checkoutRef, '') }}:
dependencySource: tag-builds
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
parameters:
dependencyList: ${{ parameters.rocmDependencies }}
gpuTarget: $(JOB_GPU_TARGET)
${{ if eq(parameters.checkoutRef, '') }}:
dependencySource: staging
${{ elseif ne(parameters.checkoutRef, '') }}:
dependencySource: tag-builds
- task: Bash@3
displayName: Add ROCm binaries to PATH
inputs:
targetType: inline
script: echo "##vso[task.prependpath]$(Agent.BuildDirectory)/rocm/bin"
- task: Bash@3
displayName: Add ROCm compilers to PATH
inputs:
targetType: inline
script: echo "##vso[task.prependpath]$(Agent.BuildDirectory)/rocm/llvm/bin"
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/build-cmake.yml
parameters:
extraBuildFlags: >-
-DCMAKE_HIP_ARCHITECTURES=$(JOB_GPU_TARGET)
-DCMAKE_C_COMPILER=$(Agent.BuildDirectory)/rocm/llvm/bin/amdclang
-DCMAKE_MODULE_PATH=$(Agent.BuildDirectory)/rocm/lib/cmake/hip
-DCMAKE_PREFIX_PATH=$(Agent.BuildDirectory)/rocm
-DCMAKE_BUILD_TYPE=Release
-DENABLE_TESTS=ON
-DINSTALL_TESTS=ON
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/gpu-diagnostics.yml
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/test.yml
parameters:
componentName: omniperf
testDir: $(Build.BinariesDirectory)/libexec/omniperf
testExecutable: export OMNIPERF_ARCH_OVERRIDE="MI300X"; ctest
- task: Bash@3
displayName: Remove ROCm binaries from PATH
inputs:
targetType: inline
script: echo "##vso[task.setvariable variable=PATH]$(echo $PATH | sed -e 's;:$(Agent.BuildDirectory)/rocm/bin;;' -e 's;^/;;' -e 's;/$;;')"
- task: Bash@3
displayName: Remove ROCm compilers from PATH
inputs:
targetType: inline
script: echo "##vso[task.setvariable variable=PATH]$(echo $PATH | sed -e 's;:$(Agent.BuildDirectory)/rocm/llvm/bin;;' -e 's;^/;;' -e 's;/$;;')"

View File

@@ -181,6 +181,7 @@ jobs:
parameters:
dependencyList: ${{ parameters.rocmDependencies }}
gpuTarget: $(JOB_GPU_TARGET)
setupHIPLibrarySymlinks: true
${{ if eq(parameters.checkoutRef, '') }}:
dependencySource: staging
${{ elseif ne(parameters.checkoutRef, '') }}:

View File

@@ -41,6 +41,7 @@ parameters:
- ROCR-Runtime
- rocprofiler-register
- roctracer
- aomp
jobs:
- job: rocprofilersdk

View File

@@ -51,6 +51,7 @@ parameters:
- rocprofiler
- rocprofiler-register
- roctracer
- rocprofiler-sdk
jobs:
- job: rocprofiler_systems
@@ -73,6 +74,12 @@ jobs:
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/checkout.yml
parameters:
checkoutRepo: ${{ parameters.checkoutRepo }}
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-aqlprofile.yml
parameters:
${{ if eq(parameters.checkoutRef, '') }}:
dependencySource: staging
${{ elseif ne(parameters.checkoutRef, '') }}:
dependencySource: tag-builds
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/dependencies-rocm.yml
parameters:
dependencyList: ${{ parameters.rocmDependencies }}
@@ -109,6 +116,7 @@ jobs:
-DROCPROFSYS_BUILD_TESTING=ON
-DROCPROFSYS_BUILD_DYNINST=ON
-DROCPROFSYS_BUILD_LIBUNWIND=ON
-DROCPROFSYS_DISABLE_EXAMPLES="openmp-target"
-DDYNINST_BUILD_TBB=ON
-DDYNINST_BUILD_ELFUTILS=ON
-DDYNINST_BUILD_LIBIBERTY=ON

View File

@@ -142,6 +142,10 @@ parameters:
- binary_ufuncs
- autograd
# - inductor/torchinductor takes too long
# set to false to disable torchvision build and test
- name: includeVision
type: boolean
default: false
trigger: none
pr: none
@@ -237,6 +241,12 @@ jobs:
git clone https://github.com/pytorch/builder.git --depth=1 --recurse-submodules
sudo ln -s $(Build.SourcesDirectory)/builder /builder
workingDirectory: $(Build.SourcesDirectory)
- task: Bash@3
displayName: Temporarily Patch CK Submodule
inputs:
targetType: inline
script: git pull origin develop
workingDirectory: $(Build.SourcesDirectory)/pytorch/third_party/composable_kernel
- task: Bash@3
displayName: Install patchelf
inputs:
@@ -296,59 +306,60 @@ jobs:
sourceDir: /remote/wheelhouserocm$(ROCM_VERSION)
contentsString: '*.whl'
# common helper source for pytorch vision and audio
- task: Bash@3
displayName: git clone pytorch test-infra
inputs:
targetType: inline
script: git clone https://github.com/pytorch/test-infra.git --depth=1 --recurse-submodules
workingDirectory: $(Build.SourcesDirectory)
- task: Bash@3
displayName: install package helper
inputs:
targetType: inline
script: python3 -m pip install test-infra/tools/pkg-helpers
workingDirectory: $(Build.SourcesDirectory)
- task: Bash@3
displayName: pytorch pkg helpers
inputs:
targetType: inline
script: CU_VERSION=${CU_VERSION} CHANNEL=${CHANNEL} python -m pytorch_pkg_helpers
# get torch vision source and build
- task: Bash@3
displayName: git clone pytorch vision
inputs:
targetType: inline
script: git clone https://github.com/pytorch/vision.git --depth=1 --recurse-submodules
workingDirectory: $(Build.SourcesDirectory)
- task: Bash@3
displayName: Build vision
inputs:
targetType: inline
script: >-
TORCH_PACKAGE_NAME=torch.$(ROCM_BRANCH).$(JOB_GPU_TARGET)
TORCHVISION_PACKAGE_NAME=torchvision.$(ROCM_BRANCH).$(JOB_GPU_TARGET)
PYTORCH_VERSION=$(cat $(Build.SourcesDirectory)/pytorch/version.txt | cut -da -f1)post$(date -u +%Y%m%d)
BUILD_VERSION=$(cat $(Build.SourcesDirectory)/vision/version.txt | cut -da -f1)post$(date -u +%Y%m%d)
python3 setup.py bdist_wheel
workingDirectory: $(Build.SourcesDirectory)/vision
- task: Bash@3
displayName: Relocate vision
inputs:
targetType: inline
script: python3 packaging/wheel/relocate.py
workingDirectory: $(Build.SourcesDirectory)/vision
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-prepare-package.yml
parameters:
sourceDir: $(Build.SourcesDirectory)/vision/dist
contentsString: '*.whl'
clean: false
- ${{ if eq(parameters.includeVision, true) }}:
- task: Bash@3
displayName: git clone pytorch test-infra
inputs:
targetType: inline
script: git clone https://github.com/pytorch/test-infra.git --depth=1 --recurse-submodules
workingDirectory: $(Build.SourcesDirectory)
- task: Bash@3
displayName: install package helper
inputs:
targetType: inline
script: python3 -m pip install test-infra/tools/pkg-helpers
workingDirectory: $(Build.SourcesDirectory)
- task: Bash@3
displayName: pytorch pkg helpers
inputs:
targetType: inline
script: CU_VERSION=${CU_VERSION} CHANNEL=${CHANNEL} python -m pytorch_pkg_helpers
# get torch vision source and build
- task: Bash@3
displayName: git clone pytorch vision
inputs:
targetType: inline
script: git clone https://github.com/pytorch/vision.git --depth=1 --recurse-submodules
workingDirectory: $(Build.SourcesDirectory)
- task: Bash@3
displayName: Build vision
inputs:
targetType: inline
script: >-
TORCH_PACKAGE_NAME=torch.$(ROCM_BRANCH).$(JOB_GPU_TARGET)
TORCHVISION_PACKAGE_NAME=torchvision.$(ROCM_BRANCH).$(JOB_GPU_TARGET)
PYTORCH_VERSION=$(cat $(Build.SourcesDirectory)/pytorch/version.txt | cut -da -f1)post$(date -u +%Y%m%d)
BUILD_VERSION=$(cat $(Build.SourcesDirectory)/vision/version.txt | cut -da -f1)post$(date -u +%Y%m%d)
python3 setup.py bdist_wheel
workingDirectory: $(Build.SourcesDirectory)/vision
- task: Bash@3
displayName: Relocate vision
inputs:
targetType: inline
script: python3 packaging/wheel/relocate.py
workingDirectory: $(Build.SourcesDirectory)/vision
- template: ${{ variables.CI_TEMPLATE_PATH }}/steps/artifact-prepare-package.yml
parameters:
sourceDir: $(Build.SourcesDirectory)/vision/dist
contentsString: '*.whl'
clean: false
- task: PublishPipelineArtifact@1
displayName: 'wheel file Publish'
retryCountOnTaskFailure: 3
inputs:
targetPath: $(Build.BinariesDirectory)
- job: torchvision_testing
- job: pytorch_testing
dependsOn: pytorch
condition: and(succeeded(), eq(variables.ENABLE_GFX942_TESTS, 'true'), not(containsValue(split(variables.DISABLED_GFX942_TESTS, ','), variables['Build.DefinitionName'])))
variables:
@@ -401,12 +412,13 @@ jobs:
targetType: inline
script: git clone https://github.com/pytorch/pytorch.git --depth=1 --recurse-submodules
workingDirectory: $(Build.SourcesDirectory)
- task: Bash@3
displayName: git clone pytorch vision
inputs:
targetType: inline
script: git clone https://github.com/pytorch/vision.git --depth=1 --recurse-submodules
workingDirectory: $(Build.SourcesDirectory)
- ${{ if eq(parameters.includeVision, true) }}:
- task: Bash@3
displayName: git clone pytorch vision
inputs:
targetType: inline
script: git clone https://github.com/pytorch/vision.git --depth=1 --recurse-submodules
workingDirectory: $(Build.SourcesDirectory)
- task: Bash@3
displayName: Install Wheel Files
inputs:
@@ -510,13 +522,14 @@ jobs:
script: pytest test/test_${{ torchTest }}.py
# Reference on what tests to run for torchvision found in private repo:
# https://github.com/ROCm/rocAutomation/blob/jenkins-pipelines/pytorch/pytorch_ci/test_torchvision.sh#L51
- task: Bash@3
displayName: Test vision/transforms
continueOnError: true
inputs:
targetType: inline
script: pytest test/test_transforms.py
workingDirectory: $(Build.SourcesDirectory)/vision
- ${{ if eq(parameters.includeVision, true) }}:
- task: Bash@3
displayName: Test vision/transforms
continueOnError: true
inputs:
targetType: inline
script: pytest test/test_transforms.py
workingDirectory: $(Build.SourcesDirectory)/vision
- task: Bash@3
displayName: Uninstall Wheel Files
inputs:

View File

@@ -26,6 +26,7 @@ parameters:
- llvm-project
- MIOpen
- MIVisionX
- omniperf
- rccl
- rdc
- rocAL

View File

@@ -0,0 +1,29 @@
variables:
- group: common
- template: /.azuredevops/variables-global.yml
parameters:
- name: checkoutRef
type: string
default: refs/tags/$(LATEST_RELEASE_TAG)
resources:
repositories:
- repository: pipelines_repo
type: github
endpoint: ROCm
name: ROCm/ROCm
- repository: release_repo
type: github
endpoint: ROCm
name: ROCm/omniperf
ref: ${{ parameters.checkoutRef }}
trigger: none
pr: none
jobs:
- template: ${{ variables.CI_COMPONENT_PATH }}/omniperf.yml
parameters:
checkoutRepo: release_repo
checkoutRef: ${{ parameters.checkoutRef }}

View File

@@ -62,7 +62,7 @@ parameters:
ROCgdb: amd-staging
rocJPEG: develop
rocm-cmake: develop
rocm-core: master
rocm-core: amd-staging
rocm-examples: develop
rocminfo: amd-staging
rocMLIR: develop

View File

@@ -165,6 +165,11 @@ parameters:
- name: skipLlvmSymlink
type: boolean
default: false
# set to true if dlopen calls for HIP libraries are causing failures
# because they do not follow shared library symlink convention
- name: setupHIPLibrarySymlinks
type: boolean
default: false
# some ROCm components can specify GPU target and this will affect downloads
- name: gpuTarget
type: string
@@ -280,6 +285,37 @@ steps:
for file in amdclang amdclang++ amdclang-cl amdclang-cpp amdflang amdlld aompcc mygpu mycpu offload-arch; do
sudo ln -s $(Agent.BuildDirectory)/rocm/llvm/bin/$file $(Agent.BuildDirectory)/rocm/bin/$file
done
# dlopen calls within a ctest or pytest sequence runs into issues when shared library symlink convention is not followed
# the convention is as follows:
# unversioned .so is a symlink to major version .so
# major version .so is a symlink to detailed version .so
# HIP libraries do not follow this convention, and each .so is a copy of each other
# changing the library structure to follow the symlink convention resolves some test failures
- ${{ if eq(parameters.setupHIPLibrarySymlinks, true) }}:
- task: Bash@3
displayName: Setup symlinks for hip libraries
inputs:
targetType: inline
workingDirectory: $(Agent.BuildDirectory)/rocm/lib
script: |
LIBRARIES=("libamdhip64" "libhiprtc-builtins" "libhiprtc")
for LIB_NAME in "${LIBRARIES[@]}"; do
VERSIONED_SO=$(ls ${LIB_NAME}.so.* 2>/dev/null | grep -E "${LIB_NAME}\.so\.[0-9]+\.[0-9]+\.[0-9]+(-.*)?" | sort -V | tail -n 1)
if [[ -z "$VERSIONED_SO" ]]; then
continue
fi
MAJOR_VERSION=$(echo "$VERSIONED_SO" | grep -oP "${LIB_NAME}\.so\.\K[0-9]+")
if [[ -e "${LIB_NAME}.so.${MAJOR_VERSION}" && ! -L "${LIB_NAME}.so.${MAJOR_VERSION}" ]]; then
rm -f "${LIB_NAME}.so.${MAJOR_VERSION}"
fi
if [[ -e "${LIB_NAME}.so" && ! -L "${LIB_NAME}.so" ]]; then
rm -f "${LIB_NAME}.so"
fi
ln -sf "$VERSIONED_SO" "${LIB_NAME}.so.${MAJOR_VERSION}"
ln -sf "${LIB_NAME}.so.${MAJOR_VERSION}" "${LIB_NAME}.so"
echo "Symlinks created for $LIB_NAME:"
ls -l ${LIB_NAME}.so*
done
- task: Bash@3
displayName: 'List downloaded ROCm files'
inputs:

View File

@@ -90,6 +90,7 @@ Dask
DataFrame
DataLoader
DataParallel
Debian
DeepSpeed
Dependabot
Deprecations
@@ -189,6 +190,7 @@ Jupyter
KFD
KFDTest
KiB
KMD
KV
KVM
Keras
@@ -211,6 +213,7 @@ MiB
MIGraphX
MIOpen
MIOpenGEMM
MIOpen's
MIVisionX
MLM
MMA
@@ -285,6 +288,7 @@ PCC
PCI
PCIe
PEFT
PEQT
PIL
PILImage
POR
@@ -314,6 +318,7 @@ RDMA
RDNA
README
RHEL
RMW
RNN
RNNs
ROC
@@ -330,6 +335,7 @@ ROCmSoftwarePlatform
ROCmValidationSuite
ROCprofiler
ROCr
RPP
RST
RW
Radeon
@@ -337,6 +343,7 @@ RelWithDebInfo
Req
Rickle
RoCE
Runfile
Ryzen
SALU
SBIOS
@@ -349,6 +356,7 @@ SENDMSG
SGPR
SGPRs
SHA
SHARK's
SIGQUIT
SIMD
SIMDs
@@ -518,6 +526,7 @@ dbgapi
de
deallocation
debuggability
debian
denoise
denoised
denoises
@@ -570,6 +579,7 @@ hipBLASLt's
hipblaslt
hipCUB
hipFFT
hipFORT
hipLIB
hipRAND
hipSOLVER
@@ -591,6 +601,7 @@ hpp
hsa
hsakmt
hyperparameter
hyperparameters
iDRAC
ib_core
inband
@@ -704,6 +715,7 @@ rocALUTION
rocBLAS
rocDecode
rocFFT
rocHPCG
rocJPEG
rocLIB
rocMLIR

View File

@@ -76,7 +76,7 @@ The Build time will reduce significantly if we limit the GPU Architecture/s agai
mkdir -p ~/WORKSPACE/ # Or any folder name other than WORKSPACE
cd ~/WORKSPACE/
export ROCM_VERSION=6.3.0
export ROCM_VERSION=6.3.1
~/bin/repo init -u http://github.com/ROCm/ROCm.git -b roc-6.3.x -m tools/rocm-build/rocm-${ROCM_VERSION}.xml
~/bin/repo sync

1530
RELEASE.md

File diff suppressed because it is too large Load Diff

View File

@@ -1,17 +1,14 @@
<?xml version="1.0" encoding="UTF-8"?>
<manifest>
<remote name="rocm-org" fetch="https://github.com/ROCm/" />
<default revision="refs/tags/rocm-6.3.0"
<default revision="refs/tags/rocm-6.3.1"
remote="rocm-org"
sync-c="true"
sync-j="4" />
<!--list of projects for ROCm-->
<project name="ROCK-Kernel-Driver" />
<project name="ROCR-Runtime" />
<project name="ROCT-Thunk-Interface" />
<project name="amdsmi" />
<project name="omniperf" />
<project name="omnitrace" />
<project name="rdc" />
<project name="rocm_bandwidth_test" />
<project name="rocm_smi_lib" />
@@ -21,6 +18,8 @@
<project name="rocprofiler" />
<project name="rocprofiler-register" />
<project name="rocprofiler-sdk" />
<project name="rocprofiler-compute" />
<project name="rocprofiler-systems" />
<project name="roctracer" />
<!--HIP Projects-->
<project name="HIP" />
@@ -42,6 +41,7 @@
<project groups="mathlibs" name="ROCmValidationSuite" />
<project groups="mathlibs" name="Tensile" />
<project groups="mathlibs" name="composable_kernel" />
<project groups="mathlibs" name="hipBLAS-common" />
<project groups="mathlibs" name="hipBLAS" />
<project groups="mathlibs" name="hipBLASLt" />
<project groups="mathlibs" name="hipCUB" />
@@ -57,6 +57,7 @@
<project groups="mathlibs" name="rocALUTION" />
<project groups="mathlibs" name="rocBLAS" />
<project groups="mathlibs" name="rocDecode" />
<project groups="mathlibs" name="rocJPEG" />
<project groups="mathlibs" name="rocPyDecode" />
<project groups="mathlibs" name="rocFFT" />
<project groups="mathlibs" name="rocPRIM" />
@@ -67,6 +68,7 @@
<project groups="mathlibs" name="rocWMMA" />
<project groups="mathlibs" name="rocm-cmake" />
<project groups="mathlibs" name="rpp" />
<project groups="mathlibs" name="TransferBench" />
<!-- Projects for OpenMP-Extras -->
<project name="aomp" path="openmp-extras/aomp" />
<project name="aomp-extras" path="openmp-extras/aomp-extras" />

View File

@@ -1,118 +1,128 @@
ROCm Version,6.3.0,6.2.4,6.2.2,6.2.1,6.2.0, 6.1.2, 6.1.1, 6.1.0, 6.0.2, 6.0.0
:ref:`Operating systems & kernels <OS-kernel-versions>`,Ubuntu 24.04.2,"Ubuntu 24.04.1, 24.04","Ubuntu 24.04.1, 24.04","Ubuntu 24.04.1, 24.04",Ubuntu 24.04,,,,,
,Ubuntu 22.04.5,"Ubuntu 22.04.5, 22.04.4","Ubuntu 22.04.5, 22.04.4","Ubuntu 22.04.5, 22.04.4","Ubuntu 22.04.5, 22.04.4","Ubuntu 22.04.4, 22.04.3","Ubuntu 22.04.4, 22.04.3","Ubuntu 22.04.4, 22.04.3","Ubuntu 22.04.4, 22.04.3, 22.04.2","Ubuntu 22.04.4, 22.04.3, 22.04.2"
,,,,,,"Ubuntu 20.04.6, 20.04.5","Ubuntu 20.04.6, 20.04.5","Ubuntu 20.04.6, 20.04.5","Ubuntu 20.04.6, 20.04.5","Ubuntu 20.04.6, 20.04.5"
,"RHEL 9.5, 9.4","RHEL 9.4, 9.3","RHEL 9.4, 9.3","RHEL 9.4, 9.3","RHEL 9.4, 9.3","RHEL 9.4 [#red-hat94-past-60]_, 9.3, 9.2","RHEL 9.4 [#red-hat94-past-60]_, 9.3, 9.2","RHEL 9.4 [#red-hat94-past-60]_, 9.3, 9.2","RHEL 9.3, 9.2","RHEL 9.3, 9.2"
,"RHEL 8.10","RHEL 8.10, 8.9","RHEL 8.10, 8.9","RHEL 8.10, 8.9","RHEL 8.10, 8.9","RHEL 8.9, 8.8","RHEL 8.9, 8.8","RHEL 8.9, 8.8","RHEL 8.9, 8.8","RHEL 8.9, 8.8"
,"SLES 15 SP6, SP5","SLES 15 SP6, SP5","SLES 15 SP6, SP5","SLES 15 SP6, SP5","SLES 15 SP6, SP5","SLES 15 SP5, SP4","SLES 15 SP5, SP4","SLES 15 SP5, SP4","SLES 15 SP5, SP4","SLES 15 SP5, SP4"
,,,,,,CentOS 7.9,CentOS 7.9,CentOS 7.9,CentOS 7.9,CentOS 7.9
,Oracle Linux 8.10 [#oracle89-past-60]_,Oracle Linux 8.9 [#oracle89-past-60]_,Oracle Linux 8.9 [#oracle89-past-60]_,Oracle Linux 8.9 [#oracle89-past-60]_,Oracle Linux 8.9 [#oracle89-past-60]_,Oracle Linux 8.9 [#oracle89-past-60]_,Oracle Linux 8.9 [#oracle89-past-60]_,,,
,.. _architecture-support-compatibility-matrix-past-60:,,,,,,,,,
:doc:`Architecture <rocm-install-on-linux:reference/system-requirements>`,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3
,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2
,CDNA,CDNA,CDNA,CDNA,CDNA,CDNA,CDNA,CDNA,CDNA,CDNA
,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3
,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2
,.. _gpu-support-compatibility-matrix-past-60:,,,,,,,,,
:doc:`GPU / LLVM target <rocm-install-on-linux:reference/system-requirements>`,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100
,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030
,gfx942,gfx942 [#mi300_624-past-60]_,gfx942 [#mi300_622-past-60]_,gfx942 [#mi300_621-past-60]_,gfx942 [#mi300_620-past-60]_, gfx942 [#mi300_612-past-60]_, gfx942 [#mi300_611-past-60]_, gfx942 [#mi300_610-past-60]_, gfx942 [#mi300_602-past-60]_, gfx942 [#mi300_600-past-60]_
,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a
,gfx908,gfx908,gfx908,gfx908,gfx908,gfx908,gfx908,gfx908,gfx908,gfx908
,,,,,,,,,,
FRAMEWORK SUPPORT,.. _framework-support-compatibility-matrix-past-60:,,,,,,,,,
:doc:`PyTorch <rocm-install-on-linux:install/3rd-party/pytorch-install>`,"2.4, 2.3, 2.2, 2.1, 2.0, 1.13","2.3, 2.2, 2.1, 2.0, 1.13","2.3, 2.2, 2.1, 2.0, 1.13","2.3, 2.2, 2.1, 2.0, 1.13","2.3, 2.2, 2.1, 2.0, 1.13","2.1, 2.0, 1.13","2.1, 2.0, 1.13","2.1, 2.0, 1.13","2.1, 2.0, 1.13","2.1, 2.0, 1.13"
:doc:`TensorFlow <rocm-install-on-linux:install/3rd-party/tensorflow-install>`,"2.17.0, 2.16.2, 2.15.1","2.16.1, 2.15.1, 2.14.1","2.16.1, 2.15.1, 2.14.1","2.16.1, 2.15.1, 2.14.1","2.16.1, 2.15.1, 2.14.1","2.15.0, 2.14.0, 2.13.1","2.15.0, 2.14.0, 2.13.1","2.15.0, 2.14.0, 2.13.1","2.14.0, 2.13.1, 2.12.1","2.14.0, 2.13.1, 2.12.1"
:doc:`JAX <rocm-install-on-linux:install/3rd-party/jax-install>`,0.4.26,0.4.26,0.4.26,0.4.26,0.4.26,0.4.26,0.4.26,0.4.26,0.4.26,0.4.26
`ONNX Runtime <https://onnxruntime.ai/docs/build/eps.html#amd-migraphx>`_,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.14.1,1.14.1
,,,,,,,,,,
THIRD PARTY COMMS,.. _thirdpartycomms-support-compatibility-matrix-past-60:,,,,,,,,,
`UCC <https://github.com/ROCm/ucc>`_,>=1.3.0,>=1.3.0,>=1.3.0,>=1.3.0,>=1.3.0,>=1.3.0,>=1.3.0,>=1.3.0,>=1.2.0,>=1.2.0
`UCX <https://github.com/ROCm/ucx>`_,>=1.15.0,>=1.15.0,>=1.15.0,>=1.15.0,>=1.15.0,>=1.14.1,>=1.14.1,>=1.14.1,>=1.14.1,>=1.14.1
,,,,,,,,,,
THIRD PARTY ALGORITHM,.. _thirdpartyalgorithm-support-compatibility-matrix-past-60:,,,,,,,,,
Thrust,2.3.2,2.2.0,2.2.0,2.2.0,2.2.0,2.1.0,2.1.0,2.1.0,2.0.1,2.0.1
CUB,2.3.2,2.2.0,2.2.0,2.2.0,2.2.0,2.1.0,2.1.0,2.1.0,2.0.1,2.0.1
,,,,,,,,,,
KFD & USER SPACE [#kfd_support-past-60]_,.. _kfd-userspace-support-compatibility-matrix-past-60:,,,,,,,,,
Tested user space versions,"6.3.x, 6.2.x, 6.1.x","6.3.x, 6.2.x, 6.1.x, 6.0.x","6.3.x, 6.2.x, 6.1.x, 6.0.x","6.3.x, 6.2.x, 6.1.x, 6.0.x","6.3.x, 6.2.x, 6.1.x, 6.0.x","6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x","6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x","6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x","6.2.x, 6.1.x, 6.0.x, 5.7.x, 5.6.x","6.2.x, 6.1.x, 6.0.x, 5.7.x, 5.6.x"
,,,,,,,,,,
ML & COMPUTER VISION,.. _mllibs-support-compatibility-matrix-past-60:,,,,,,,,,
:doc:`Composable Kernel <composable_kernel:index>`,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0
:doc:`MIGraphX <amdmigraphx:index>`,2.11.0,2.10.0,2.10.0,2.10.0,2.10.0,2.9.0,2.9.0,2.9.0,2.8.0,2.8.0
:doc:`MIOpen <miopen:index>`,3.3.0,3.2.0,3.2.0,3.2.0,3.2.0,3.1.0,3.1.0,3.1.0,3.0.0,3.0.0
:doc:`MIVisionX <mivisionx:index>`,3.1.0,3.0.0,3.0.0,3.0.0,3.0.0,2.5.0,2.5.0,2.5.0,2.5.0,2.5.0
:doc:`rocAL <rocal:index>`,2.1.0,2.0.0,2.0.0,2.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0
:doc:`rocDecode <rocdecode:index>`,0.8.0,0.6.0,0.6.0,0.6.0,0.6.0,0.6.0,0.5.0,0.5.0,N/A,N/A
:doc:`rocJPEG <rocjpeg:index>`,0.6.0,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A
:doc:`rocPyDecode <rocpydecode:index>`,0.2.0,0.1.0,0.1.0,0.1.0,0.1.0,N/A,N/A,N/A,N/A,N/A
:doc:`RPP <rpp:index>`,1.9.1,1.8.0,1.8.0,1.8.0,1.8.0,1.5.0,1.5.0,1.5.0,1.4.0,1.4.0
,,,,,,,,,,
COMMUNICATION,.. _commlibs-support-compatibility-matrix-past-60:,,,,,,,,,
:doc:`RCCL <rccl:index>`,2.21.5,2.20.5,2.20.5,2.20.5,2.20.5,2.18.6,2.18.6,2.18.6,2.18.3,2.18.3
,,,,,,,,,,
MATH LIBS,.. _mathlibs-support-compatibility-matrix-past-60:,,,,,,,,,
`half <https://github.com/ROCm/half>`_ ,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0
:doc:`hipBLAS <hipblas:index>`,2.3.0,2.2.0,2.2.0,2.2.0,2.2.0,2.1.0,2.1.0,2.1.0,2.0.0,2.0.0
:doc:`hipBLASLt <hipblaslt:index>`,0.10.0,0.8.0,0.8.0,0.8.0,0.8.0,0.7.0,0.7.0,0.7.0,0.6.0,0.6.0
:doc:`hipFFT <hipfft:index>`,1.0.17,1.0.16,1.0.15,1.0.15,1.0.14,1.0.14,1.0.14,1.0.14,1.0.13,1.0.13
:doc:`hipfort <hipfort:index>`,0.5.0,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0
:doc:`hipRAND <hiprand:index>`,2.11.0,2.11.1,2.11.0,2.11.0,2.11.0,2.10.16,2.10.16,2.10.16,2.10.16,2.10.16
:doc:`hipSOLVER <hipsolver:index>`,2.3.0,2.2.0,2.2.0,2.2.0,2.2.0,2.1.1,2.1.1,2.1.0,2.0.0,2.0.0
:doc:`hipSPARSE <hipsparse:index>`,3.1.2,3.1.1,3.1.1,3.1.1,3.1.1,3.0.1,3.0.1,3.0.1,3.0.0,3.0.0
:doc:`hipSPARSELt <hipsparselt:index>`,0.2.2,0.2.1,0.2.1,0.2.1,0.2.1,0.2.0,0.1.0,0.1.0,0.1.0,0.1.0
:doc:`rocALUTION <rocalution:index>`,3.2.1,3.2.1,3.2.0,3.2.0,3.2.0,3.1.1,3.1.1,3.1.1,3.0.3,3.0.3
:doc:`rocBLAS <rocblas:index>`,4.3.0,4.2.4,4.2.1,4.2.1,4.2.0,4.1.2,4.1.0,4.1.0,4.0.0,4.0.0
:doc:`rocFFT <rocfft:index>`,1.0.31,1.0.30,1.0.29,1.0.29,1.0.28,1.0.27,1.0.27,1.0.26,1.0.25,1.0.23
:doc:`rocRAND <rocrand:index>`,3.2.0,3.1.1,3.1.0,3.1.0,3.1.0,3.0.1,3.0.1,3.0.1,3.0.0,2.10.17
:doc:`rocSOLVER <rocsolver:index>`,3.27.0,3.26.2,3.26.0,3.26.0,3.26.0,3.25.0,3.25.0,3.25.0,3.24.0,3.24.0
:doc:`rocSPARSE <rocsparse:index>`,3.3.0,3.2.1,3.2.0,3.2.0,3.2.0,3.1.2,3.1.2,3.1.2,3.0.2,3.0.2
:doc:`rocWMMA <rocwmma:index>`,1.6.0,1.5.0,1.5.0,1.5.0,1.5.0,1.4.0,1.4.0,1.4.0,1.3.0,1.3.0
:doc:`Tensile <tensile:index>`,4.42.0,4.41.0,4.41.0,4.41.0,4.41.0,4.40.0,4.40.0,4.40.0,4.39.0,4.39.0
,,,,,,,,,,
PRIMITIVES,.. _primitivelibs-support-compatibility-matrix-past-60:,,,,,,,,,
:doc:`hipCUB <hipcub:index>`,3.3.0,3.2.1,3.2.0,3.2.0,3.2.0,3.1.0,3.1.0,3.1.0,3.0.0,3.0.0
:doc:`hipTensor <hiptensor:index>`,1.4.0,1.3.0,1.3.0,1.3.0,1.3.0,1.2.0,1.2.0,1.2.0,1.1.0,1.1.0
:doc:`rocPRIM <rocprim:index>`,3.3.0,3.2.2,3.2.0,3.2.0,3.2.0,3.1.0,3.1.0,3.1.0,3.0.0,3.0.0
:doc:`rocThrust <rocthrust:index>`,3.3.0,3.1.1,3.1.0,3.1.0,3.0.1,3.0.1,3.0.1,3.0.1,3.0.0,3.0.0
,,,,,,,,,,
SUPPORT LIBS,,,,,,,,,,
`hipother <https://github.com/ROCm/hipother>`_,6.3.42131,6.2.41134,6.2.41134,6.2.41134,6.2.41133,6.1.40093,6.1.40092,6.1.40091,6.1.32831,6.1.32830
`rocm-core <https://github.com/ROCm/rocm-core>`_,6.3.0,6.2.4,6.2.2,6.2.1,6.2.0,6.1.2,6.1.1,6.1.0,6.0.2,6.0.0
`ROCT-Thunk-Interface <https://github.com/ROCm/ROCT-Thunk-Interface>`_,N/A [#ROCT-rocr-past-60]_,20240607.5.7,20240607.5.7,20240607.4.05,20240607.1.4246,20240125.5.08,20240125.5.08,20240125.3.30,20231016.2.245,20231016.2.245
,,,,,,,,,,
SYSTEM MGMT TOOLS,.. _tools-support-compatibility-matrix-past-60:,,,,,,,,,
:doc:`AMD SMI <amdsmi:index>`,24.7.1,24.6.3,24.6.3,24.6.3,24.6.2,24.5.1,24.5.1,24.4.1,23.4.2,23.4.2
:doc:`ROCm Data Center Tool <rdc:index>`,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0
:doc:`rocminfo <rocminfo:index>`,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0
:doc:`ROCm SMI <rocm_smi_lib:index>`,7.4.0,7.3.0,7.3.0,7.3.0,7.3.0,7.2.0,7.0.0,7.0.0,6.0.2,6.0.0
:doc:`ROCm Validation Suite <rocmvalidationsuite:index>`,1.1.0,1.0.60204,1.0.60202,1.0.60201,1.0.60200,1.0.60102,1.0.60101,1.0.60100,1.0.60002,1.0.60000
,,,,,,,,,,
PERFORMANCE TOOLS,,,,,,,,,,
:doc:`ROCm Bandwidth Test <rocm_bandwidth_test:index>`,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0
:doc:`ROCm Compute Profiler <rocprofiler-compute:index>`,3.0.0,2.0.1,2.0.1,2.0.1,2.0.1,N/A,N/A,N/A,N/A,N/A
:doc:`ROCm Systems Profiler <rocprofiler-systems:index>`,0.1.0,1.11.2,1.11.2,1.11.2,1.11.2,N/A,N/A,N/A,N/A,N/A
:doc:`ROCProfiler <rocprofiler:index>`,2.0.60300,2.0.60204,2.0.60202,2.0.60201,2.0.60200,2.0.60102,2.0.60101,2.0.60100,2.0.60002,2.0.60000
:doc:`ROCprofiler-SDK <rocprofiler-sdk:index>`,0.5.0,0.4.0,0.4.0,0.4.0,0.4.0,N/A,N/A,N/A,N/A,N/A
:doc:`ROCTracer <roctracer:index>`,4.1.60300,4.1.60204,4.1.60202,4.1.60201,4.1.60200,4.1.60102,4.1.60101,4.1.60100,4.1.60002,4.1.60000
,,,,,,,,,,
DEVELOPMENT TOOLS,,,,,,,,,,
:doc:`HIPIFY <hipify:index>`,18.0.0.24455,18.0.0.24392,18.0.0.24355,18.0.0.24355,18.0.0.24232,17.0.0.24193,17.0.0.24154,17.0.0.24103,17.0.0.24012,17.0.0.23483
:doc:`ROCm CMake <rocmcmakebuildtools:index>`,0.14.0,0.13.0,0.13.0,0.13.0,0.13.0,0.12.0,0.12.0,0.12.0,0.11.0,0.11.0
:doc:`ROCdbgapi <rocdbgapi:index>`,0.77.0,0.76.0,0.76.0,0.76.0,0.76.0,0.71.0,0.71.0,0.71.0,0.71.0,0.71.0
:doc:`ROCm Debugger (ROCgdb) <rocgdb:index>`,15.2.0,14.2.0,14.2.0,14.2.0,14.2.0,14.1.0,14.1.0,14.1.0,13.2.0,13.2.0
`rocprofiler-register <https://github.com/ROCm/rocprofiler-register>`_,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0,0.3.0,0.3.0,0.3.0,N/A,N/A
:doc:`ROCr Debug Agent <rocr_debug_agent:index>`,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3
,,,,,,,,,,
COMPILERS,.. _compilers-support-compatibility-matrix-past-60:,,,,,,,,,
`clang-ocl <https://github.com/ROCm/clang-ocl>`_,N/A,N/A,N/A,N/A,N/A,0.5.0,0.5.0,0.5.0,0.5.0,0.5.0
:doc:`hipCC <hipcc:index>`,1.1.1,1.1.1,1.1.1,1.1.1,1.1.1,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0
`Flang <https://github.com/ROCm/flang>`_,18.0.0.24455,18.0.0.24392,18.0.0.24355,18.0.0.24355,18.0.0.24232,17.0.0.24193,17.0.0.24154,17.0.0.24103,17.0.0.24012,17.0.0.23483
:doc:`llvm-project <llvm-project:index>`,18.0.0.24455,18.0.0.24392,18.0.0.24355,18.0.0.24355,18.0.0.24232,17.0.0.24193,17.0.0.24154,17.0.0.24103,17.0.0.24012,17.0.0.23483
`OpenMP <https://github.com/ROCm/llvm-project/tree/amd-staging/openmp>`_,18.0.0.24455,18.0.0.24392,18.0.0.24355,18.0.0.24355,18.0.0.24232,17.0.0.24193,17.0.0.24154,17.0.0.24103,17.0.0.24012,17.0.0.23483
,,,,,,,,,,
RUNTIMES,.. _runtime-support-compatibility-matrix-past-60:,,,,,,,,,
:doc:`AMD CLR <hip:understand/amd_clr>`,6.3.42131,6.2.41134,6.2.41134,6.2.41134,6.2.41133,6.1.40093,6.1.40092,6.1.40091,6.1.32831,6.1.32830
:doc:`HIP <hip:index>`,6.3.42131,6.2.41134,6.2.41134,6.2.41134,6.2.41133,6.1.40093,6.1.40092,6.1.40091,6.1.32831,6.1.32830
`OpenCL Runtime <https://github.com/ROCm/clr/tree/develop/opencl>`_,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0
:doc:`ROCr Runtime <rocr-runtime:index>`,1.14.0,1.14.0,1.14.0,1.14.0,1.13.0,1.13.0,1.13.0,1.13.0,1.12.0,1.12.0
ROCm Version,6.3.1,6.3.0,6.2.4,6.2.2,6.2.1,6.2.0, 6.1.2, 6.1.1, 6.1.0, 6.0.2, 6.0.0
:ref:`Operating systems & kernels <OS-kernel-versions>`,Ubuntu 24.04.2,Ubuntu 24.04.2,"Ubuntu 24.04.1, 24.04","Ubuntu 24.04.1, 24.04","Ubuntu 24.04.1, 24.04",Ubuntu 24.04,,,,,
,Ubuntu 22.04.5,Ubuntu 22.04.5,"Ubuntu 22.04.5, 22.04.4","Ubuntu 22.04.5, 22.04.4","Ubuntu 22.04.5, 22.04.4","Ubuntu 22.04.5, 22.04.4","Ubuntu 22.04.4, 22.04.3","Ubuntu 22.04.4, 22.04.3","Ubuntu 22.04.4, 22.04.3","Ubuntu 22.04.4, 22.04.3, 22.04.2","Ubuntu 22.04.4, 22.04.3, 22.04.2"
,,,,,,,"Ubuntu 20.04.6, 20.04.5","Ubuntu 20.04.6, 20.04.5","Ubuntu 20.04.6, 20.04.5","Ubuntu 20.04.6, 20.04.5","Ubuntu 20.04.6, 20.04.5"
,"RHEL 9.5, 9.4","RHEL 9.5, 9.4","RHEL 9.4, 9.3","RHEL 9.4, 9.3","RHEL 9.4, 9.3","RHEL 9.4, 9.3","RHEL 9.4, 9.3, 9.2","RHEL 9.4, 9.3, 9.2","RHEL 9.4, 9.3, 9.2","RHEL 9.3, 9.2","RHEL 9.3, 9.2"
,RHEL 8.10,RHEL 8.10,"RHEL 8.10, 8.9","RHEL 8.10, 8.9","RHEL 8.10, 8.9","RHEL 8.10, 8.9","RHEL 8.9, 8.8","RHEL 8.9, 8.8","RHEL 8.9, 8.8","RHEL 8.9, 8.8","RHEL 8.9, 8.8"
,"SLES 15 SP6, SP5","SLES 15 SP6, SP5","SLES 15 SP6, SP5","SLES 15 SP6, SP5","SLES 15 SP6, SP5","SLES 15 SP6, SP5","SLES 15 SP5, SP4","SLES 15 SP5, SP4","SLES 15 SP5, SP4","SLES 15 SP5, SP4","SLES 15 SP5, SP4"
,,,,,,,CentOS 7.9,CentOS 7.9,CentOS 7.9,CentOS 7.9,CentOS 7.9
,Oracle Linux 8.10 [#mic300x-past-60]_,Oracle Linux 8.10 [#mic300x-past-60]_,Oracle Linux 8.9 [#mic300x-past-60]_,Oracle Linux 8.9 [#mic300x-past-60]_,Oracle Linux 8.9 [#mic300x-past-60]_,Oracle Linux 8.9 [#mic300x-past-60]_,Oracle Linux 8.9 [#mic300x-past-60]_,Oracle Linux 8.9 [#mic300x-past-60]_,,,
,Debian 12 [#mic300x-past-60]_,,,,,,,,,,
,.. _architecture-support-compatibility-matrix-past-60:,,,,,,,,,,
:doc:`Architecture <rocm-install-on-linux:reference/system-requirements>`,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3,CDNA3
,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2,CDNA2
,CDNA,CDNA,CDNA,CDNA,CDNA,CDNA,CDNA,CDNA,CDNA,CDNA,CDNA
,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3,RDNA3
,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2,RDNA2
,.. _gpu-support-compatibility-matrix-past-60:,,,,,,,,,,
:doc:`GPU / LLVM target <rocm-install-on-linux:reference/system-requirements>`,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100,gfx1100
,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030,gfx1030
,gfx942,gfx942,gfx942 [#mi300_624-past-60]_,gfx942 [#mi300_622-past-60]_,gfx942 [#mi300_621-past-60]_,gfx942 [#mi300_620-past-60]_, gfx942 [#mi300_612-past-60]_, gfx942 [#mi300_611-past-60]_, gfx942 [#mi300_610-past-60]_, gfx942 [#mi300_602-past-60]_, gfx942 [#mi300_600-past-60]_
,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a,gfx90a
,gfx908,gfx908,gfx908,gfx908,gfx908,gfx908,gfx908,gfx908,gfx908,gfx908,gfx908
,,,,,,,,,,,
FRAMEWORK SUPPORT,.. _framework-support-compatibility-matrix-past-60:,,,,,,,,,,
:doc:`PyTorch <rocm-install-on-linux:install/3rd-party/pytorch-install>`,"2.4, 2.3, 2.2, 2.1, 2.0, 1.13","2.4, 2.3, 2.2, 2.1, 2.0, 1.13","2.3, 2.2, 2.1, 2.0, 1.13","2.3, 2.2, 2.1, 2.0, 1.13","2.3, 2.2, 2.1, 2.0, 1.13","2.3, 2.2, 2.1, 2.0, 1.13","2.1, 2.0, 1.13","2.1, 2.0, 1.13","2.1, 2.0, 1.13","2.1, 2.0, 1.13","2.1, 2.0, 1.13"
:doc:`TensorFlow <rocm-install-on-linux:install/3rd-party/tensorflow-install>`,"2.17.0, 2.16.2, 2.15.1","2.17.0, 2.16.2, 2.15.1","2.16.1, 2.15.1, 2.14.1","2.16.1, 2.15.1, 2.14.1","2.16.1, 2.15.1, 2.14.1","2.16.1, 2.15.1, 2.14.1","2.15.0, 2.14.0, 2.13.1","2.15.0, 2.14.0, 2.13.1","2.15.0, 2.14.0, 2.13.1","2.14.0, 2.13.1, 2.12.1","2.14.0, 2.13.1, 2.12.1"
:doc:`JAX <rocm-install-on-linux:install/3rd-party/jax-install>`,0.4.35,0.4.35,0.4.26,0.4.26,0.4.26,0.4.26,0.4.26,0.4.26,0.4.26,0.4.26,0.4.26
`ONNX Runtime <https://onnxruntime.ai/docs/build/eps.html#amd-migraphx>`_,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.14.1,1.14.1
,,,,,,,,,,,
THIRD PARTY COMMS,.. _thirdpartycomms-support-compatibility-matrix-past-60:,,,,,,,,,,
`UCC <https://github.com/ROCm/ucc>`_,>=1.3.0,>=1.3.0,>=1.3.0,>=1.3.0,>=1.3.0,>=1.3.0,>=1.3.0,>=1.3.0,>=1.3.0,>=1.2.0,>=1.2.0
`UCX <https://github.com/ROCm/ucx>`_,>=1.15.0,>=1.15.0,>=1.15.0,>=1.15.0,>=1.15.0,>=1.15.0,>=1.14.1,>=1.14.1,>=1.14.1,>=1.14.1,>=1.14.1
,,,,,,,,,,,
THIRD PARTY ALGORITHM,.. _thirdpartyalgorithm-support-compatibility-matrix-past-60:,,,,,,,,,,
Thrust,2.3.2,2.3.2,2.2.0,2.2.0,2.2.0,2.2.0,2.1.0,2.1.0,2.1.0,2.0.1,2.0.1
CUB,2.3.2,2.3.2,2.2.0,2.2.0,2.2.0,2.2.0,2.1.0,2.1.0,2.1.0,2.0.1,2.0.1
,,,,,,,,,,,
,,,,,,,,,,,
KMD & USER SPACE [#kfd_support-past-60]_,.. _kfd-userspace-support-compatibility-matrix-past-60:,,,,,,,,,
Tested user space versions,"6.3.x, 6.2.x, 6.1.x","6.3.x, 6.2.x, 6.1.x","6.3.x, 6.2.x, 6.1.x, 6.0.x","6.3.x, 6.2.x, 6.1.x, 6.0.x","6.3.x, 6.2.x, 6.1.x, 6.0.x","6.3.x, 6.2.x, 6.1.x, 6.0.x","6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x","6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x","6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x","6.2.x, 6.1.x, 6.0.x, 5.7.x, 5.6.x","6.2.x, 6.1.x, 6.0.x, 5.7.x, 5.6.x"
,,,,,,,,,,,
ML & COMPUTER VISION,.. _mllibs-support-compatibility-matrix-past-60:,,,,,,,,,,
:doc:`Composable Kernel <composable_kernel:index>`,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0,1.1.0
:doc:`MIGraphX <amdmigraphx:index>`,2.11.0,2.11.0,2.10.0,2.10.0,2.10.0,2.10.0,2.9.0,2.9.0,2.9.0,2.8.0,2.8.0
:doc:`MIOpen <miopen:index>`,3.3.0,3.3.0,3.2.0,3.2.0,3.2.0,3.2.0,3.1.0,3.1.0,3.1.0,3.0.0,3.0.0
:doc:`MIVisionX <mivisionx:index>`,3.1.0,3.1.0,3.0.0,3.0.0,3.0.0,3.0.0,2.5.0,2.5.0,2.5.0,2.5.0,2.5.0
:doc:`rocAL <rocal:index>`,2.1.0,2.1.0,2.0.0,2.0.0,2.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0
:doc:`rocDecode <rocdecode:index>`,0.8.0,0.8.0,0.6.0,0.6.0,0.6.0,0.6.0,0.6.0,0.5.0,0.5.0,N/A,N/A
:doc:`rocJPEG <rocjpeg:index>`,0.6.0,0.6.0,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A
:doc:`rocPyDecode <rocpydecode:index>`,0.2.0,0.2.0,0.1.0,0.1.0,0.1.0,0.1.0,N/A,N/A,N/A,N/A,N/A
:doc:`RPP <rpp:index>`,1.9.1,1.9.1,1.8.0,1.8.0,1.8.0,1.8.0,1.5.0,1.5.0,1.5.0,1.4.0,1.4.0
,,,,,,,,,,,
COMMUNICATION,.. _commlibs-support-compatibility-matrix-past-60:,,,,,,,,,,
:doc:`RCCL <rccl:index>`,2.21.5,2.21.5,2.20.5,2.20.5,2.20.5,2.20.5,2.18.6,2.18.6,2.18.6,2.18.3,2.18.3
,,,,,,,,,,,
MATH LIBS,.. _mathlibs-support-compatibility-matrix-past-60:,,,,,,,,,,
`half <https://github.com/ROCm/half>`_ ,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0,1.12.0
:doc:`hipBLAS <hipblas:index>`,2.3.0,2.3.0,2.2.0,2.2.0,2.2.0,2.2.0,2.1.0,2.1.0,2.1.0,2.0.0,2.0.0
:doc:`hipBLASLt <hipblaslt:index>`,0.10.0,0.10.0,0.8.0,0.8.0,0.8.0,0.8.0,0.7.0,0.7.0,0.7.0,0.6.0,0.6.0
:doc:`hipFFT <hipfft:index>`,1.0.17,1.0.17,1.0.16,1.0.15,1.0.15,1.0.14,1.0.14,1.0.14,1.0.14,1.0.13,1.0.13
:doc:`hipfort <hipfort:index>`,0.5.0,0.5.0,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0
:doc:`hipRAND <hiprand:index>`,2.11.1,2.11.0,2.11.1,2.11.0,2.11.0,2.11.0,2.10.16,2.10.16,2.10.16,2.10.16,2.10.16
,,,,,,,,,,,
:doc:`hipSOLVER <hipsolver:index>`,2.3.0,2.3.0,2.2.0,2.2.0,2.2.0,2.2.0,2.1.1,2.1.1,2.1.0,2.0.0,2.0.0
:doc:`hipSPARSE <hipsparse:index>`,3.1.2,3.1.2,3.1.1,3.1.1,3.1.1,3.1.1,3.0.1,3.0.1,3.0.1,3.0.0,3.0.0
:doc:`hipSPARSELt <hipsparselt:index>`,0.2.2,0.2.2,0.2.1,0.2.1,0.2.1,0.2.1,0.2.0,0.1.0,0.1.0,0.1.0,0.1.0
:doc:`rocALUTION <rocalution:index>`,3.2.1,3.2.1,3.2.1,3.2.0,3.2.0,3.2.0,3.1.1,3.1.1,3.1.1,3.0.3,3.0.3
:doc:`rocBLAS <rocblas:index>`,4.3.0,4.3.0,4.2.4,4.2.1,4.2.1,4.2.0,4.1.2,4.1.0,4.1.0,4.0.0,4.0.0
:doc:`rocFFT <rocfft:index>`,1.0.31,1.0.31,1.0.30,1.0.29,1.0.29,1.0.28,1.0.27,1.0.27,1.0.26,1.0.25,1.0.23
:doc:`rocRAND <rocrand:index>`,3.2.0,3.2.0,3.1.1,3.1.0,3.1.0,3.1.0,3.0.1,3.0.1,3.0.1,3.0.0,2.10.17
:doc:`rocSOLVER <rocsolver:index>`,3.27.0,3.27.0,3.26.2,3.26.0,3.26.0,3.26.0,3.25.0,3.25.0,3.25.0,3.24.0,3.24.0
:doc:`rocSPARSE <rocsparse:index>`,3.3.0,3.3.0,3.2.1,3.2.0,3.2.0,3.2.0,3.1.2,3.1.2,3.1.2,3.0.2,3.0.2
:doc:`rocWMMA <rocwmma:index>`,1.6.0,1.6.0,1.5.0,1.5.0,1.5.0,1.5.0,1.4.0,1.4.0,1.4.0,1.3.0,1.3.0
:doc:`Tensile <tensile:index>`,4.42.0,4.42.0,4.41.0,4.41.0,4.41.0,4.41.0,4.40.0,4.40.0,4.40.0,4.39.0,4.39.0
,,,,,,,,,,,
PRIMITIVES,.. _primitivelibs-support-compatibility-matrix-past-60:,,,,,,,,,,
:doc:`hipCUB <hipcub:index>`,3.3.0,3.3.0,3.2.1,3.2.0,3.2.0,3.2.0,3.1.0,3.1.0,3.1.0,3.0.0,3.0.0
:doc:`hipTensor <hiptensor:index>`,1.4.0,1.4.0,1.3.0,1.3.0,1.3.0,1.3.0,1.2.0,1.2.0,1.2.0,1.1.0,1.1.0
:doc:`rocPRIM <rocprim:index>`,3.3.0,3.3.0,3.2.2,3.2.0,3.2.0,3.2.0,3.1.0,3.1.0,3.1.0,3.0.0,3.0.0
:doc:`rocThrust <rocthrust:index>`,3.3.0,3.3.0,3.1.1,3.1.0,3.1.0,3.0.1,3.0.1,3.0.1,3.0.1,3.0.0,3.0.0
,,,,,,,,,,,
SUPPORT LIBS,,,,,,,,,,,
`hipother <https://github.com/ROCm/hipother>`_,6.3.42133,6.3.42131,6.2.41134,6.2.41134,6.2.41134,6.2.41133,6.1.40093,6.1.40092,6.1.40091,6.1.32831,6.1.32830
,,,,,,,,,,,
`rocm-core <https://github.com/ROCm/rocm-core>`_,6.3.1,6.3.0,6.2.4,6.2.2,6.2.1,6.2.0,6.1.2,6.1.1,6.1.0,6.0.2,6.0.0
`ROCT-Thunk-Interface <https://github.com/ROCm/ROCT-Thunk-Interface>`_,N/A [#ROCT-rocr-past-60]_,N/A [#ROCT-rocr-past-60]_,20240607.5.7,20240607.5.7,20240607.4.05,20240607.1.4246,20240125.5.08,20240125.5.08,20240125.3.30,20231016.2.245,20231016.2.245
,,,,,,,,,,,
SYSTEM MGMT TOOLS,.. _tools-support-compatibility-matrix-past-60:,,,,,,,,,,
:doc:`AMD SMI <amdsmi:index>`,24.7.1,24.7.1,24.6.3,24.6.3,24.6.3,24.6.2,24.5.1,24.5.1,24.4.1,23.4.2,23.4.2
:doc:`ROCm Data Center Tool <rdc:index>`,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0,0.3.0
:doc:`rocminfo <rocminfo:index>`,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0
:doc:`ROCm SMI <rocm_smi_lib:index>`,7.4.0,7.4.0,7.3.0,7.3.0,7.3.0,7.3.0,7.2.0,7.0.0,7.0.0,6.0.2,6.0.0
:doc:`ROCm Validation Suite <rocmvalidationsuite:index>`,1.1.0,1.1.0,1.0.60204,1.0.60202,1.0.60201,1.0.60200,1.0.60102,1.0.60101,1.0.60100,1.0.60002,1.0.60000
,,,,,,,,,,,
PERFORMANCE TOOLS,,,,,,,,,,,
:doc:`ROCm Bandwidth Test <rocm_bandwidth_test:index>`,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0,1.4.0
:doc:`ROCm Compute Profiler <rocprofiler-compute:index>`,3.0.0,3.0.0,2.0.1,2.0.1,2.0.1,2.0.1,N/A,N/A,N/A,N/A,N/A
:doc:`ROCm Systems Profiler <rocprofiler-systems:index>`,0.1.0,0.1.0,1.11.2,1.11.2,1.11.2,1.11.2,N/A,N/A,N/A,N/A,N/A
:doc:`ROCProfiler <rocprofiler:index>`,2.0.60301,2.0.60300,2.0.60204,2.0.60202,2.0.60201,2.0.60200,2.0.60102,2.0.60101,2.0.60100,2.0.60002,2.0.60000
,,,,,,,,,,,
:doc:`ROCprofiler-SDK <rocprofiler-sdk:index>`,0.5.0,0.5.0,0.4.0,0.4.0,0.4.0,0.4.0,N/A,N/A,N/A,N/A,N/A
:doc:`ROCTracer <roctracer:index>`,4.1.60301,4.1.60300,4.1.60204,4.1.60202,4.1.60201,4.1.60200,4.1.60102,4.1.60101,4.1.60100,4.1.60002,4.1.60000
,,,,,,,,,,,
,,,,,,,,,,,
DEVELOPMENT TOOLS,,,,,,,,,,,
:doc:`HIPIFY <hipify:index>`,18.0.0.24491,18.0.0.24455,18.0.0.24392,18.0.0.24355,18.0.0.24355,18.0.0.24232,17.0.0.24193,17.0.0.24154,17.0.0.24103,17.0.0.24012,17.0.0.23483
,,,,,,,,,,,
:doc:`ROCm CMake <rocmcmakebuildtools:index>`,0.14.0,0.14.0,0.13.0,0.13.0,0.13.0,0.13.0,0.12.0,0.12.0,0.12.0,0.11.0,0.11.0
:doc:`ROCdbgapi <rocdbgapi:index>`,0.77.0,0.77.0,0.76.0,0.76.0,0.76.0,0.76.0,0.71.0,0.71.0,0.71.0,0.71.0,0.71.0
:doc:`ROCm Debugger (ROCgdb) <rocgdb:index>`,15.2.0,15.2.0,14.2.0,14.2.0,14.2.0,14.2.0,14.1.0,14.1.0,14.1.0,13.2.0,13.2.0
`rocprofiler-register <https://github.com/ROCm/rocprofiler-register>`_,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0,0.4.0,0.3.0,0.3.0,0.3.0,N/A,N/A
:doc:`ROCr Debug Agent <rocr_debug_agent:index>`,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3,2.0.3
,,,,,,,,,,,
COMPILERS,.. _compilers-support-compatibility-matrix-past-60:,,,,,,,,,,
`clang-ocl <https://github.com/ROCm/clang-ocl>`_,N/A,N/A,N/A,N/A,N/A,N/A,0.5.0,0.5.0,0.5.0,0.5.0,0.5.0
:doc:`hipCC <hipcc:index>`,1.1.1,1.1.1,1.1.1,1.1.1,1.1.1,1.1.1,1.0.0,1.0.0,1.0.0,1.0.0,1.0.0
`Flang <https://github.com/ROCm/flang>`_,18.0.0.24491,18.0.0.24455,18.0.0.24392,18.0.0.24355,18.0.0.24355,18.0.0.24232,17.0.0.24193,17.0.0.24154,17.0.0.24103,17.0.0.24012,17.0.0.23483
:doc:`llvm-project <llvm-project:index>`,18.0.0.24455,18.0.0.24491,18.0.0.24392,18.0.0.24355,18.0.0.24355,18.0.0.24232,17.0.0.24193,17.0.0.24154,17.0.0.24103,17.0.0.24012,17.0.0.23483
`OpenMP <https://github.com/ROCm/llvm-project/tree/amd-staging/openmp>`_,18.0.0.24455,18.0.0.24491,18.0.0.24392,18.0.0.24355,18.0.0.24355,18.0.0.24232,17.0.0.24193,17.0.0.24154,17.0.0.24103,17.0.0.24012,17.0.0.23483
,,,,,,,,,,,
,,,,,,,,,,,
RUNTIMES,.. _runtime-support-compatibility-matrix-past-60:,,,,,,,,,,
:doc:`AMD CLR <hip:understand/amd_clr>`,6.3.42133,6.3.42131,6.2.41134,6.2.41134,6.2.41134,6.2.41133,6.1.40093,6.1.40092,6.1.40091,6.1.32831,6.1.32830
,,,,,,,,,,,
:doc:`HIP <hip:index>`,6.3.42133,6.3.42131,6.2.41134,6.2.41134,6.2.41134,6.2.41133,6.1.40093,6.1.40092,6.1.40091,6.1.32831,6.1.32830
,,,,,,,,,,,
`OpenCL Runtime <https://github.com/ROCm/clr/tree/develop/opencl>`_,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0,2.0.0
:doc:`ROCr Runtime <rocr-runtime:index>`,1.14.0,1.14.0,1.14.0,1.14.0,1.14.0,1.13.0,1.13.0,1.13.0,1.13.0,1.12.0,1.12.0
1 ROCm Version 6.3.1 6.3.0 6.2.4 6.2.2 6.2.1 6.2.0 6.1.2 6.1.1 6.1.0 6.0.2 6.0.0
2 :ref:`Operating systems & kernels <OS-kernel-versions>` Ubuntu 24.04.2 Ubuntu 24.04.2 Ubuntu 24.04.1, 24.04 Ubuntu 24.04.1, 24.04 Ubuntu 24.04.1, 24.04 Ubuntu 24.04
3 Ubuntu 22.04.5 Ubuntu 22.04.5 Ubuntu 22.04.5, 22.04.4 Ubuntu 22.04.5, 22.04.4 Ubuntu 22.04.5, 22.04.4 Ubuntu 22.04.5, 22.04.4 Ubuntu 22.04.4, 22.04.3 Ubuntu 22.04.4, 22.04.3 Ubuntu 22.04.4, 22.04.3 Ubuntu 22.04.4, 22.04.3, 22.04.2 Ubuntu 22.04.4, 22.04.3, 22.04.2
4 Ubuntu 20.04.6, 20.04.5 Ubuntu 20.04.6, 20.04.5 Ubuntu 20.04.6, 20.04.5 Ubuntu 20.04.6, 20.04.5 Ubuntu 20.04.6, 20.04.5
5 RHEL 9.5, 9.4 RHEL 9.5, 9.4 RHEL 9.4, 9.3 RHEL 9.4, 9.3 RHEL 9.4, 9.3 RHEL 9.4, 9.3 RHEL 9.4 [#red-hat94-past-60]_, 9.3, 9.2 RHEL 9.4, 9.3, 9.2 RHEL 9.4 [#red-hat94-past-60]_, 9.3, 9.2 RHEL 9.4, 9.3, 9.2 RHEL 9.4 [#red-hat94-past-60]_, 9.3, 9.2 RHEL 9.4, 9.3, 9.2 RHEL 9.3, 9.2 RHEL 9.3, 9.2
6 RHEL 8.10 RHEL 8.10 RHEL 8.10, 8.9 RHEL 8.10, 8.9 RHEL 8.10, 8.9 RHEL 8.10, 8.9 RHEL 8.9, 8.8 RHEL 8.9, 8.8 RHEL 8.9, 8.8 RHEL 8.9, 8.8 RHEL 8.9, 8.8
7 SLES 15 SP6, SP5 SLES 15 SP6, SP5 SLES 15 SP6, SP5 SLES 15 SP6, SP5 SLES 15 SP6, SP5 SLES 15 SP6, SP5 SLES 15 SP5, SP4 SLES 15 SP5, SP4 SLES 15 SP5, SP4 SLES 15 SP5, SP4 SLES 15 SP5, SP4
8 CentOS 7.9 CentOS 7.9 CentOS 7.9 CentOS 7.9 CentOS 7.9
9 Oracle Linux 8.10 [#mic300x-past-60]_ Oracle Linux 8.10 [#oracle89-past-60]_ Oracle Linux 8.10 [#mic300x-past-60]_ Oracle Linux 8.9 [#oracle89-past-60]_ Oracle Linux 8.9 [#mic300x-past-60]_ Oracle Linux 8.9 [#oracle89-past-60]_ Oracle Linux 8.9 [#mic300x-past-60]_ Oracle Linux 8.9 [#oracle89-past-60]_ Oracle Linux 8.9 [#mic300x-past-60]_ Oracle Linux 8.9 [#oracle89-past-60]_ Oracle Linux 8.9 [#mic300x-past-60]_ Oracle Linux 8.9 [#oracle89-past-60]_ Oracle Linux 8.9 [#mic300x-past-60]_ Oracle Linux 8.9 [#oracle89-past-60]_ Oracle Linux 8.9 [#mic300x-past-60]_
10 Debian 12 [#mic300x-past-60]_ .. _architecture-support-compatibility-matrix-past-60:
11 :doc:`Architecture <rocm-install-on-linux:reference/system-requirements>` .. _architecture-support-compatibility-matrix-past-60: CDNA3 CDNA3 CDNA3 CDNA3 CDNA3 CDNA3 CDNA3 CDNA3 CDNA3 CDNA3
12 :doc:`Architecture <rocm-install-on-linux:reference/system-requirements>` CDNA3 CDNA2 CDNA3 CDNA2 CDNA3 CDNA2 CDNA3 CDNA2 CDNA3 CDNA2 CDNA3 CDNA2 CDNA3 CDNA2 CDNA3 CDNA2 CDNA3 CDNA2 CDNA3 CDNA2 CDNA3
13 CDNA2 CDNA CDNA2 CDNA CDNA2 CDNA CDNA2 CDNA CDNA2 CDNA CDNA2 CDNA CDNA2 CDNA CDNA2 CDNA CDNA2 CDNA CDNA2 CDNA CDNA2
14 CDNA RDNA3 CDNA RDNA3 CDNA RDNA3 CDNA RDNA3 CDNA RDNA3 CDNA RDNA3 CDNA RDNA3 CDNA RDNA3 CDNA RDNA3 CDNA RDNA3 CDNA
15 RDNA3 RDNA2 RDNA3 RDNA2 RDNA3 RDNA2 RDNA3 RDNA2 RDNA3 RDNA2 RDNA3 RDNA2 RDNA3 RDNA2 RDNA3 RDNA2 RDNA3 RDNA2 RDNA3 RDNA2 RDNA3
16 RDNA2 .. _gpu-support-compatibility-matrix-past-60: RDNA2 RDNA2 RDNA2 RDNA2 RDNA2 RDNA2 RDNA2 RDNA2 RDNA2 RDNA2
17 :doc:`GPU / LLVM target <rocm-install-on-linux:reference/system-requirements>` .. _gpu-support-compatibility-matrix-past-60: gfx1100 gfx1100 gfx1100 gfx1100 gfx1100 gfx1100 gfx1100 gfx1100 gfx1100 gfx1100
18 :doc:`GPU / LLVM target <rocm-install-on-linux:reference/system-requirements>` gfx1100 gfx1030 gfx1100 gfx1030 gfx1100 gfx1030 gfx1100 gfx1030 gfx1100 gfx1030 gfx1100 gfx1030 gfx1100 gfx1030 gfx1100 gfx1030 gfx1100 gfx1030 gfx1100 gfx1030 gfx1100
19 gfx1030 gfx942 gfx1030 gfx942 [#mi300_624-past-60]_ gfx1030 gfx942 [#mi300_622-past-60]_ gfx1030 gfx942 [#mi300_621-past-60]_ gfx1030 gfx942 [#mi300_620-past-60]_ gfx1030 gfx942 [#mi300_612-past-60]_ gfx1030 gfx942 [#mi300_611-past-60]_ gfx1030 gfx942 [#mi300_610-past-60]_ gfx1030 gfx942 [#mi300_602-past-60]_ gfx1030 gfx942 [#mi300_600-past-60]_ gfx1030
20 gfx942 gfx90a gfx942 gfx90a gfx942 [#mi300_624-past-60]_ gfx90a gfx942 [#mi300_622-past-60]_ gfx90a gfx942 [#mi300_621-past-60]_ gfx90a gfx942 [#mi300_620-past-60]_ gfx90a gfx942 [#mi300_612-past-60]_ gfx90a gfx942 [#mi300_611-past-60]_ gfx90a gfx942 [#mi300_610-past-60]_ gfx90a gfx942 [#mi300_602-past-60]_ gfx90a gfx942 [#mi300_600-past-60]_
21 gfx90a gfx908 gfx90a gfx908 gfx90a gfx908 gfx90a gfx908 gfx90a gfx908 gfx90a gfx908 gfx90a gfx908 gfx90a gfx908 gfx90a gfx908 gfx90a gfx908 gfx90a
22 gfx908 gfx908 gfx908 gfx908 gfx908 gfx908 gfx908 gfx908 gfx908 gfx908 gfx908
23 FRAMEWORK SUPPORT .. _framework-support-compatibility-matrix-past-60:
24 :doc:`PyTorch <rocm-install-on-linux:install/3rd-party/pytorch-install>` FRAMEWORK SUPPORT .. _framework-support-compatibility-matrix-past-60: 2.4, 2.3, 2.2, 2.1, 2.0, 1.13 2.3, 2.2, 2.1, 2.0, 1.13 2.3, 2.2, 2.1, 2.0, 1.13 2.3, 2.2, 2.1, 2.0, 1.13 2.3, 2.2, 2.1, 2.0, 1.13 2.1, 2.0, 1.13 2.1, 2.0, 1.13 2.1, 2.0, 1.13 2.1, 2.0, 1.13 2.1, 2.0, 1.13
25 :doc:`TensorFlow <rocm-install-on-linux:install/3rd-party/tensorflow-install>` :doc:`PyTorch <rocm-install-on-linux:install/3rd-party/pytorch-install>` 2.4, 2.3, 2.2, 2.1, 2.0, 1.13 2.17.0, 2.16.2, 2.15.1 2.4, 2.3, 2.2, 2.1, 2.0, 1.13 2.16.1, 2.15.1, 2.14.1 2.3, 2.2, 2.1, 2.0, 1.13 2.16.1, 2.15.1, 2.14.1 2.3, 2.2, 2.1, 2.0, 1.13 2.16.1, 2.15.1, 2.14.1 2.3, 2.2, 2.1, 2.0, 1.13 2.16.1, 2.15.1, 2.14.1 2.3, 2.2, 2.1, 2.0, 1.13 2.15.0, 2.14.0, 2.13.1 2.1, 2.0, 1.13 2.15.0, 2.14.0, 2.13.1 2.1, 2.0, 1.13 2.15.0, 2.14.0, 2.13.1 2.1, 2.0, 1.13 2.14.0, 2.13.1, 2.12.1 2.1, 2.0, 1.13 2.14.0, 2.13.1, 2.12.1 2.1, 2.0, 1.13
26 :doc:`JAX <rocm-install-on-linux:install/3rd-party/jax-install>` :doc:`TensorFlow <rocm-install-on-linux:install/3rd-party/tensorflow-install>` 2.17.0, 2.16.2, 2.15.1 0.4.26 2.17.0, 2.16.2, 2.15.1 0.4.26 2.16.1, 2.15.1, 2.14.1 0.4.26 2.16.1, 2.15.1, 2.14.1 0.4.26 2.16.1, 2.15.1, 2.14.1 0.4.26 2.16.1, 2.15.1, 2.14.1 0.4.26 2.15.0, 2.14.0, 2.13.1 0.4.26 2.15.0, 2.14.0, 2.13.1 0.4.26 2.15.0, 2.14.0, 2.13.1 0.4.26 2.14.0, 2.13.1, 2.12.1 0.4.26 2.14.0, 2.13.1, 2.12.1
27 `ONNX Runtime <https://onnxruntime.ai/docs/build/eps.html#amd-migraphx>`_ :doc:`JAX <rocm-install-on-linux:install/3rd-party/jax-install>` 0.4.35 1.17.3 0.4.35 1.17.3 0.4.26 1.17.3 0.4.26 1.17.3 0.4.26 1.17.3 0.4.26 1.17.3 0.4.26 1.17.3 0.4.26 1.17.3 0.4.26 1.14.1 0.4.26 1.14.1 0.4.26
28 `ONNX Runtime <https://onnxruntime.ai/docs/build/eps.html#amd-migraphx>`_ 1.17.3 1.17.3 1.17.3 1.17.3 1.17.3 1.17.3 1.17.3 1.17.3 1.17.3 1.14.1 1.14.1
29 THIRD PARTY COMMS .. _thirdpartycomms-support-compatibility-matrix-past-60:
30 `UCC <https://github.com/ROCm/ucc>`_ THIRD PARTY COMMS .. _thirdpartycomms-support-compatibility-matrix-past-60: >=1.3.0 >=1.3.0 >=1.3.0 >=1.3.0 >=1.3.0 >=1.3.0 >=1.3.0 >=1.3.0 >=1.2.0 >=1.2.0
31 `UCX <https://github.com/ROCm/ucx>`_ `UCC <https://github.com/ROCm/ucc>`_ >=1.3.0 >=1.15.0 >=1.3.0 >=1.15.0 >=1.3.0 >=1.15.0 >=1.3.0 >=1.15.0 >=1.3.0 >=1.15.0 >=1.3.0 >=1.14.1 >=1.3.0 >=1.14.1 >=1.3.0 >=1.14.1 >=1.3.0 >=1.14.1 >=1.2.0 >=1.14.1 >=1.2.0
32 `UCX <https://github.com/ROCm/ucx>`_ >=1.15.0 >=1.15.0 >=1.15.0 >=1.15.0 >=1.15.0 >=1.15.0 >=1.14.1 >=1.14.1 >=1.14.1 >=1.14.1 >=1.14.1
33 THIRD PARTY ALGORITHM .. _thirdpartyalgorithm-support-compatibility-matrix-past-60:
34 Thrust THIRD PARTY ALGORITHM .. _thirdpartyalgorithm-support-compatibility-matrix-past-60: 2.3.2 2.2.0 2.2.0 2.2.0 2.2.0 2.1.0 2.1.0 2.1.0 2.0.1 2.0.1
35 CUB Thrust 2.3.2 2.3.2 2.2.0 2.2.0 2.2.0 2.2.0 2.1.0 2.1.0 2.1.0 2.0.1 2.0.1
36 CUB 2.3.2 2.3.2 2.2.0 2.2.0 2.2.0 2.2.0 2.1.0 2.1.0 2.1.0 2.0.1 2.0.1
37 KFD & USER SPACE [#kfd_support-past-60]_ .. _kfd-userspace-support-compatibility-matrix-past-60:
38 Tested user space versions 6.3.x, 6.2.x, 6.1.x 6.3.x, 6.2.x, 6.1.x, 6.0.x 6.3.x, 6.2.x, 6.1.x, 6.0.x 6.3.x, 6.2.x, 6.1.x, 6.0.x 6.3.x, 6.2.x, 6.1.x, 6.0.x 6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x 6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x 6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x 6.2.x, 6.1.x, 6.0.x, 5.7.x, 5.6.x 6.2.x, 6.1.x, 6.0.x, 5.7.x, 5.6.x
39 KMD & USER SPACE [#kfd_support-past-60]_ .. _kfd-userspace-support-compatibility-matrix-past-60:
40 ML & COMPUTER VISION Tested user space versions 6.3.x, 6.2.x, 6.1.x .. _mllibs-support-compatibility-matrix-past-60: 6.3.x, 6.2.x, 6.1.x 6.3.x, 6.2.x, 6.1.x, 6.0.x 6.3.x, 6.2.x, 6.1.x, 6.0.x 6.3.x, 6.2.x, 6.1.x, 6.0.x 6.3.x, 6.2.x, 6.1.x, 6.0.x 6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x 6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x 6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x 6.2.x, 6.1.x, 6.0.x, 5.7.x, 5.6.x 6.2.x, 6.1.x, 6.0.x, 5.7.x, 5.6.x
41 :doc:`Composable Kernel <composable_kernel:index>` 1.1.0 1.1.0 1.1.0 1.1.0 1.1.0 1.1.0 1.1.0 1.1.0 1.1.0 1.1.0
42 :doc:`MIGraphX <amdmigraphx:index>` ML & COMPUTER VISION .. _mllibs-support-compatibility-matrix-past-60: 2.11.0 2.10.0 2.10.0 2.10.0 2.10.0 2.9.0 2.9.0 2.9.0 2.8.0 2.8.0
43 :doc:`MIOpen <miopen:index>` :doc:`Composable Kernel <composable_kernel:index>` 1.1.0 3.3.0 1.1.0 3.2.0 1.1.0 3.2.0 1.1.0 3.2.0 1.1.0 3.2.0 1.1.0 3.1.0 1.1.0 3.1.0 1.1.0 3.1.0 1.1.0 3.0.0 1.1.0 3.0.0 1.1.0
44 :doc:`MIVisionX <mivisionx:index>` :doc:`MIGraphX <amdmigraphx:index>` 2.11.0 3.1.0 2.11.0 3.0.0 2.10.0 3.0.0 2.10.0 3.0.0 2.10.0 3.0.0 2.10.0 2.5.0 2.9.0 2.5.0 2.9.0 2.5.0 2.9.0 2.5.0 2.8.0 2.5.0 2.8.0
45 :doc:`rocAL <rocal:index>` :doc:`MIOpen <miopen:index>` 3.3.0 2.1.0 3.3.0 2.0.0 3.2.0 2.0.0 3.2.0 2.0.0 3.2.0 1.0.0 3.2.0 1.0.0 3.1.0 1.0.0 3.1.0 1.0.0 3.1.0 1.0.0 3.0.0 1.0.0 3.0.0
46 :doc:`rocDecode <rocdecode:index>` :doc:`MIVisionX <mivisionx:index>` 3.1.0 0.8.0 3.1.0 0.6.0 3.0.0 0.6.0 3.0.0 0.6.0 3.0.0 0.6.0 3.0.0 0.6.0 2.5.0 0.5.0 2.5.0 0.5.0 2.5.0 N/A 2.5.0 N/A 2.5.0
47 :doc:`rocJPEG <rocjpeg:index>` :doc:`rocAL <rocal:index>` 2.1.0 0.6.0 2.1.0 N/A 2.0.0 N/A 2.0.0 N/A 2.0.0 N/A 1.0.0 N/A 1.0.0 N/A 1.0.0 N/A 1.0.0 N/A 1.0.0 N/A 1.0.0
48 :doc:`rocPyDecode <rocpydecode:index>` :doc:`rocDecode <rocdecode:index>` 0.8.0 0.2.0 0.8.0 0.1.0 0.6.0 0.1.0 0.6.0 0.1.0 0.6.0 0.1.0 0.6.0 N/A 0.6.0 N/A 0.5.0 N/A 0.5.0 N/A N/A
49 :doc:`RPP <rpp:index>` :doc:`rocJPEG <rocjpeg:index>` 0.6.0 1.9.1 0.6.0 1.8.0 N/A 1.8.0 N/A 1.8.0 N/A 1.8.0 N/A 1.5.0 N/A 1.5.0 N/A 1.5.0 N/A 1.4.0 N/A 1.4.0 N/A
50 :doc:`rocPyDecode <rocpydecode:index>` 0.2.0 0.2.0 0.1.0 0.1.0 0.1.0 0.1.0 N/A N/A N/A N/A N/A
51 COMMUNICATION :doc:`RPP <rpp:index>` 1.9.1 .. _commlibs-support-compatibility-matrix-past-60: 1.9.1 1.8.0 1.8.0 1.8.0 1.8.0 1.5.0 1.5.0 1.5.0 1.4.0 1.4.0
52 :doc:`RCCL <rccl:index>` 2.21.5 2.20.5 2.20.5 2.20.5 2.20.5 2.18.6 2.18.6 2.18.6 2.18.3 2.18.3
53 COMMUNICATION .. _commlibs-support-compatibility-matrix-past-60:
54 MATH LIBS :doc:`RCCL <rccl:index>` 2.21.5 .. _mathlibs-support-compatibility-matrix-past-60: 2.21.5 2.20.5 2.20.5 2.20.5 2.20.5 2.18.6 2.18.6 2.18.6 2.18.3 2.18.3
55 `half <https://github.com/ROCm/half>`_ 1.12.0 1.12.0 1.12.0 1.12.0 1.12.0 1.12.0 1.12.0 1.12.0 1.12.0 1.12.0
56 :doc:`hipBLAS <hipblas:index>` MATH LIBS .. _mathlibs-support-compatibility-matrix-past-60: 2.3.0 2.2.0 2.2.0 2.2.0 2.2.0 2.1.0 2.1.0 2.1.0 2.0.0 2.0.0
57 :doc:`hipBLASLt <hipblaslt:index>` `half <https://github.com/ROCm/half>`_ 1.12.0 0.10.0 1.12.0 0.8.0 1.12.0 0.8.0 1.12.0 0.8.0 1.12.0 0.8.0 1.12.0 0.7.0 1.12.0 0.7.0 1.12.0 0.7.0 1.12.0 0.6.0 1.12.0 0.6.0 1.12.0
58 :doc:`hipFFT <hipfft:index>` :doc:`hipBLAS <hipblas:index>` 2.3.0 1.0.17 2.3.0 1.0.16 2.2.0 1.0.15 2.2.0 1.0.15 2.2.0 1.0.14 2.2.0 1.0.14 2.1.0 1.0.14 2.1.0 1.0.14 2.1.0 1.0.13 2.0.0 1.0.13 2.0.0
59 :doc:`hipfort <hipfort:index>` :doc:`hipBLASLt <hipblaslt:index>` 0.10.0 0.5.0 0.10.0 0.4.0 0.8.0 0.4.0 0.8.0 0.4.0 0.8.0 0.4.0 0.8.0 0.4.0 0.7.0 0.4.0 0.7.0 0.4.0 0.7.0 0.4.0 0.6.0 0.4.0 0.6.0
60 :doc:`hipRAND <hiprand:index>` :doc:`hipFFT <hipfft:index>` 1.0.17 2.11.0 1.0.17 2.11.1 1.0.16 2.11.0 1.0.15 2.11.0 1.0.15 2.11.0 1.0.14 2.10.16 1.0.14 2.10.16 1.0.14 2.10.16 1.0.14 2.10.16 1.0.13 2.10.16 1.0.13
61 :doc:`hipSOLVER <hipsolver:index>` :doc:`hipfort <hipfort:index>` 0.5.0 2.3.0 0.5.0 2.2.0 0.4.0 2.2.0 0.4.0 2.2.0 0.4.0 2.2.0 0.4.0 2.1.1 0.4.0 2.1.1 0.4.0 2.1.0 0.4.0 2.0.0 0.4.0 2.0.0 0.4.0
62 :doc:`hipSPARSE <hipsparse:index>` :doc:`hipRAND <hiprand:index>` 2.11.1 3.1.2 2.11.0 3.1.1 2.11.1 3.1.1 2.11.0 3.1.1 2.11.0 3.1.1 2.11.0 3.0.1 2.10.16 3.0.1 2.10.16 3.0.1 2.10.16 3.0.0 2.10.16 3.0.0 2.10.16
63 :doc:`hipSPARSELt <hipsparselt:index>` 0.2.2 0.2.1 0.2.1 0.2.1 0.2.1 0.2.0 0.1.0 0.1.0 0.1.0 0.1.0
64 :doc:`rocALUTION <rocalution:index>` :doc:`hipSOLVER <hipsolver:index>` 2.3.0 3.2.1 2.3.0 3.2.1 2.2.0 3.2.0 2.2.0 3.2.0 2.2.0 3.2.0 2.2.0 3.1.1 2.1.1 3.1.1 2.1.1 3.1.1 2.1.0 3.0.3 2.0.0 3.0.3 2.0.0
65 :doc:`rocBLAS <rocblas:index>` :doc:`hipSPARSE <hipsparse:index>` 3.1.2 4.3.0 3.1.2 4.2.4 3.1.1 4.2.1 3.1.1 4.2.1 3.1.1 4.2.0 3.1.1 4.1.2 3.0.1 4.1.0 3.0.1 4.1.0 3.0.1 4.0.0 3.0.0 4.0.0 3.0.0
66 :doc:`rocFFT <rocfft:index>` :doc:`hipSPARSELt <hipsparselt:index>` 0.2.2 1.0.31 0.2.2 1.0.30 0.2.1 1.0.29 0.2.1 1.0.29 0.2.1 1.0.28 0.2.1 1.0.27 0.2.0 1.0.27 0.1.0 1.0.26 0.1.0 1.0.25 0.1.0 1.0.23 0.1.0
67 :doc:`rocRAND <rocrand:index>` :doc:`rocALUTION <rocalution:index>` 3.2.1 3.2.0 3.2.1 3.1.1 3.2.1 3.1.0 3.2.0 3.1.0 3.2.0 3.1.0 3.2.0 3.0.1 3.1.1 3.0.1 3.1.1 3.0.1 3.1.1 3.0.0 3.0.3 2.10.17 3.0.3
68 :doc:`rocSOLVER <rocsolver:index>` :doc:`rocBLAS <rocblas:index>` 4.3.0 3.27.0 4.3.0 3.26.2 4.2.4 3.26.0 4.2.1 3.26.0 4.2.1 3.26.0 4.2.0 3.25.0 4.1.2 3.25.0 4.1.0 3.25.0 4.1.0 3.24.0 4.0.0 3.24.0 4.0.0
69 :doc:`rocSPARSE <rocsparse:index>` :doc:`rocFFT <rocfft:index>` 1.0.31 3.3.0 1.0.31 3.2.1 1.0.30 3.2.0 1.0.29 3.2.0 1.0.29 3.2.0 1.0.28 3.1.2 1.0.27 3.1.2 1.0.27 3.1.2 1.0.26 3.0.2 1.0.25 3.0.2 1.0.23
70 :doc:`rocWMMA <rocwmma:index>` :doc:`rocRAND <rocrand:index>` 3.2.0 1.6.0 3.2.0 1.5.0 3.1.1 1.5.0 3.1.0 1.5.0 3.1.0 1.5.0 3.1.0 1.4.0 3.0.1 1.4.0 3.0.1 1.4.0 3.0.1 1.3.0 3.0.0 1.3.0 2.10.17
71 :doc:`Tensile <tensile:index>` :doc:`rocSOLVER <rocsolver:index>` 3.27.0 4.42.0 3.27.0 4.41.0 3.26.2 4.41.0 3.26.0 4.41.0 3.26.0 4.41.0 3.26.0 4.40.0 3.25.0 4.40.0 3.25.0 4.40.0 3.25.0 4.39.0 3.24.0 4.39.0 3.24.0
72 :doc:`rocSPARSE <rocsparse:index>` 3.3.0 3.3.0 3.2.1 3.2.0 3.2.0 3.2.0 3.1.2 3.1.2 3.1.2 3.0.2 3.0.2
73 PRIMITIVES :doc:`rocWMMA <rocwmma:index>` 1.6.0 .. _primitivelibs-support-compatibility-matrix-past-60: 1.6.0 1.5.0 1.5.0 1.5.0 1.5.0 1.4.0 1.4.0 1.4.0 1.3.0 1.3.0
74 :doc:`hipCUB <hipcub:index>` :doc:`Tensile <tensile:index>` 4.42.0 3.3.0 4.42.0 3.2.1 4.41.0 3.2.0 4.41.0 3.2.0 4.41.0 3.2.0 4.41.0 3.1.0 4.40.0 3.1.0 4.40.0 3.1.0 4.40.0 3.0.0 4.39.0 3.0.0 4.39.0
75 :doc:`hipTensor <hiptensor:index>` 1.4.0 1.3.0 1.3.0 1.3.0 1.3.0 1.2.0 1.2.0 1.2.0 1.1.0 1.1.0
76 :doc:`rocPRIM <rocprim:index>` PRIMITIVES .. _primitivelibs-support-compatibility-matrix-past-60: 3.3.0 3.2.2 3.2.0 3.2.0 3.2.0 3.1.0 3.1.0 3.1.0 3.0.0 3.0.0
77 :doc:`rocThrust <rocthrust:index>` :doc:`hipCUB <hipcub:index>` 3.3.0 3.3.0 3.1.1 3.2.1 3.1.0 3.2.0 3.1.0 3.2.0 3.0.1 3.2.0 3.0.1 3.1.0 3.0.1 3.1.0 3.0.1 3.1.0 3.0.0 3.0.0
78 :doc:`hipTensor <hiptensor:index>` 1.4.0 1.4.0 1.3.0 1.3.0 1.3.0 1.3.0 1.2.0 1.2.0 1.2.0 1.1.0 1.1.0
79 SUPPORT LIBS :doc:`rocPRIM <rocprim:index>` 3.3.0 3.3.0 3.2.2 3.2.0 3.2.0 3.2.0 3.1.0 3.1.0 3.1.0 3.0.0 3.0.0
80 `hipother <https://github.com/ROCm/hipother>`_ :doc:`rocThrust <rocthrust:index>` 3.3.0 6.3.42131 3.3.0 6.2.41134 3.1.1 6.2.41134 3.1.0 6.2.41134 3.1.0 6.2.41133 3.0.1 6.1.40093 3.0.1 6.1.40092 3.0.1 6.1.40091 3.0.1 6.1.32831 3.0.0 6.1.32830 3.0.0
81 `rocm-core <https://github.com/ROCm/rocm-core>`_ 6.3.0 6.2.4 6.2.2 6.2.1 6.2.0 6.1.2 6.1.1 6.1.0 6.0.2 6.0.0
82 `ROCT-Thunk-Interface <https://github.com/ROCm/ROCT-Thunk-Interface>`_ SUPPORT LIBS N/A [#ROCT-rocr-past-60]_ 20240607.5.7 20240607.5.7 20240607.4.05 20240607.1.4246 20240125.5.08 20240125.5.08 20240125.3.30 20231016.2.245 20231016.2.245
83 `hipother <https://github.com/ROCm/hipother>`_ 6.3.42133 6.3.42131 6.2.41134 6.2.41134 6.2.41134 6.2.41133 6.1.40093 6.1.40092 6.1.40091 6.1.32831 6.1.32830
84 SYSTEM MGMT TOOLS .. _tools-support-compatibility-matrix-past-60:
85 :doc:`AMD SMI <amdsmi:index>` `rocm-core <https://github.com/ROCm/rocm-core>`_ 6.3.1 24.7.1 6.3.0 24.6.3 6.2.4 24.6.3 6.2.2 24.6.3 6.2.1 24.6.2 6.2.0 24.5.1 6.1.2 24.5.1 6.1.1 24.4.1 6.1.0 23.4.2 6.0.2 23.4.2 6.0.0
86 :doc:`ROCm Data Center Tool <rdc:index>` `ROCT-Thunk-Interface <https://github.com/ROCm/ROCT-Thunk-Interface>`_ N/A [#ROCT-rocr-past-60]_ 0.3.0 N/A [#ROCT-rocr-past-60]_ 0.3.0 20240607.5.7 0.3.0 20240607.5.7 0.3.0 20240607.4.05 0.3.0 20240607.1.4246 0.3.0 20240125.5.08 0.3.0 20240125.5.08 0.3.0 20240125.3.30 0.3.0 20231016.2.245 0.3.0 20231016.2.245
87 :doc:`rocminfo <rocminfo:index>` 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0
88 :doc:`ROCm SMI <rocm_smi_lib:index>` SYSTEM MGMT TOOLS .. _tools-support-compatibility-matrix-past-60: 7.4.0 7.3.0 7.3.0 7.3.0 7.3.0 7.2.0 7.0.0 7.0.0 6.0.2 6.0.0
89 :doc:`ROCm Validation Suite <rocmvalidationsuite:index>` :doc:`AMD SMI <amdsmi:index>` 24.7.1 1.1.0 24.7.1 1.0.60204 24.6.3 1.0.60202 24.6.3 1.0.60201 24.6.3 1.0.60200 24.6.2 1.0.60102 24.5.1 1.0.60101 24.5.1 1.0.60100 24.4.1 1.0.60002 23.4.2 1.0.60000 23.4.2
90 :doc:`ROCm Data Center Tool <rdc:index>` 0.3.0 0.3.0 0.3.0 0.3.0 0.3.0 0.3.0 0.3.0 0.3.0 0.3.0 0.3.0 0.3.0
91 PERFORMANCE TOOLS :doc:`rocminfo <rocminfo:index>` 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0 1.0.0
92 :doc:`ROCm Bandwidth Test <rocm_bandwidth_test:index>` :doc:`ROCm SMI <rocm_smi_lib:index>` 7.4.0 1.4.0 7.4.0 1.4.0 7.3.0 1.4.0 7.3.0 1.4.0 7.3.0 1.4.0 7.3.0 1.4.0 7.2.0 1.4.0 7.0.0 1.4.0 7.0.0 1.4.0 6.0.2 1.4.0 6.0.0
93 :doc:`ROCm Compute Profiler <rocprofiler-compute:index>` :doc:`ROCm Validation Suite <rocmvalidationsuite:index>` 1.1.0 3.0.0 1.1.0 2.0.1 1.0.60204 2.0.1 1.0.60202 2.0.1 1.0.60201 2.0.1 1.0.60200 N/A 1.0.60102 N/A 1.0.60101 N/A 1.0.60100 N/A 1.0.60002 N/A 1.0.60000
94 :doc:`ROCm Systems Profiler <rocprofiler-systems:index>` 0.1.0 1.11.2 1.11.2 1.11.2 1.11.2 N/A N/A N/A N/A N/A
95 :doc:`ROCProfiler <rocprofiler:index>` PERFORMANCE TOOLS 2.0.60300 2.0.60204 2.0.60202 2.0.60201 2.0.60200 2.0.60102 2.0.60101 2.0.60100 2.0.60002 2.0.60000
96 :doc:`ROCprofiler-SDK <rocprofiler-sdk:index>` :doc:`ROCm Bandwidth Test <rocm_bandwidth_test:index>` 1.4.0 0.5.0 1.4.0 0.4.0 1.4.0 0.4.0 1.4.0 0.4.0 1.4.0 0.4.0 1.4.0 N/A 1.4.0 N/A 1.4.0 N/A 1.4.0 N/A 1.4.0 N/A 1.4.0
97 :doc:`ROCTracer <roctracer:index>` :doc:`ROCm Compute Profiler <rocprofiler-compute:index>` 3.0.0 4.1.60300 3.0.0 4.1.60204 2.0.1 4.1.60202 2.0.1 4.1.60201 2.0.1 4.1.60200 2.0.1 4.1.60102 N/A 4.1.60101 N/A 4.1.60100 N/A 4.1.60002 N/A 4.1.60000 N/A
98 :doc:`ROCm Systems Profiler <rocprofiler-systems:index>` 0.1.0 0.1.0 1.11.2 1.11.2 1.11.2 1.11.2 N/A N/A N/A N/A N/A
99 DEVELOPMENT TOOLS :doc:`ROCProfiler <rocprofiler:index>` 2.0.60301 2.0.60300 2.0.60204 2.0.60202 2.0.60201 2.0.60200 2.0.60102 2.0.60101 2.0.60100 2.0.60002 2.0.60000
100 :doc:`HIPIFY <hipify:index>` 18.0.0.24455 18.0.0.24392 18.0.0.24355 18.0.0.24355 18.0.0.24232 17.0.0.24193 17.0.0.24154 17.0.0.24103 17.0.0.24012 17.0.0.23483
101 :doc:`ROCm CMake <rocmcmakebuildtools:index>` :doc:`ROCprofiler-SDK <rocprofiler-sdk:index>` 0.5.0 0.14.0 0.5.0 0.13.0 0.4.0 0.13.0 0.4.0 0.13.0 0.4.0 0.13.0 0.4.0 0.12.0 N/A 0.12.0 N/A 0.12.0 N/A 0.11.0 N/A 0.11.0 N/A
102 :doc:`ROCdbgapi <rocdbgapi:index>` :doc:`ROCTracer <roctracer:index>` 4.1.60301 0.77.0 4.1.60300 0.76.0 4.1.60204 0.76.0 4.1.60202 0.76.0 4.1.60201 0.76.0 4.1.60200 0.71.0 4.1.60102 0.71.0 4.1.60101 0.71.0 4.1.60100 0.71.0 4.1.60002 0.71.0 4.1.60000
103 :doc:`ROCm Debugger (ROCgdb) <rocgdb:index>` 15.2.0 14.2.0 14.2.0 14.2.0 14.2.0 14.1.0 14.1.0 14.1.0 13.2.0 13.2.0
104 `rocprofiler-register <https://github.com/ROCm/rocprofiler-register>`_ 0.4.0 0.4.0 0.4.0 0.4.0 0.4.0 0.3.0 0.3.0 0.3.0 N/A N/A
105 :doc:`ROCr Debug Agent <rocr_debug_agent:index>` DEVELOPMENT TOOLS 2.0.3 2.0.3 2.0.3 2.0.3 2.0.3 2.0.3 2.0.3 2.0.3 2.0.3 2.0.3
106 :doc:`HIPIFY <hipify:index>` 18.0.0.24491 18.0.0.24455 18.0.0.24392 18.0.0.24355 18.0.0.24355 18.0.0.24232 17.0.0.24193 17.0.0.24154 17.0.0.24103 17.0.0.24012 17.0.0.23483
107 COMPILERS .. _compilers-support-compatibility-matrix-past-60:
108 `clang-ocl <https://github.com/ROCm/clang-ocl>`_ :doc:`ROCm CMake <rocmcmakebuildtools:index>` 0.14.0 N/A 0.14.0 N/A 0.13.0 N/A 0.13.0 N/A 0.13.0 N/A 0.13.0 0.5.0 0.12.0 0.5.0 0.12.0 0.5.0 0.12.0 0.5.0 0.11.0 0.5.0 0.11.0
109 :doc:`hipCC <hipcc:index>` :doc:`ROCdbgapi <rocdbgapi:index>` 0.77.0 1.1.1 0.77.0 1.1.1 0.76.0 1.1.1 0.76.0 1.1.1 0.76.0 1.1.1 0.76.0 1.0.0 0.71.0 1.0.0 0.71.0 1.0.0 0.71.0 1.0.0 0.71.0 1.0.0 0.71.0
110 `Flang <https://github.com/ROCm/flang>`_ :doc:`ROCm Debugger (ROCgdb) <rocgdb:index>` 15.2.0 18.0.0.24455 15.2.0 18.0.0.24392 14.2.0 18.0.0.24355 14.2.0 18.0.0.24355 14.2.0 18.0.0.24232 14.2.0 17.0.0.24193 14.1.0 17.0.0.24154 14.1.0 17.0.0.24103 14.1.0 17.0.0.24012 13.2.0 17.0.0.23483 13.2.0
111 :doc:`llvm-project <llvm-project:index>` `rocprofiler-register <https://github.com/ROCm/rocprofiler-register>`_ 0.4.0 18.0.0.24455 0.4.0 18.0.0.24392 0.4.0 18.0.0.24355 0.4.0 18.0.0.24355 0.4.0 18.0.0.24232 0.4.0 17.0.0.24193 0.3.0 17.0.0.24154 0.3.0 17.0.0.24103 0.3.0 17.0.0.24012 N/A 17.0.0.23483 N/A
112 `OpenMP <https://github.com/ROCm/llvm-project/tree/amd-staging/openmp>`_ :doc:`ROCr Debug Agent <rocr_debug_agent:index>` 2.0.3 18.0.0.24455 2.0.3 18.0.0.24392 2.0.3 18.0.0.24355 2.0.3 18.0.0.24355 2.0.3 18.0.0.24232 2.0.3 17.0.0.24193 2.0.3 17.0.0.24154 2.0.3 17.0.0.24103 2.0.3 17.0.0.24012 2.0.3 17.0.0.23483 2.0.3
113
114 RUNTIMES COMPILERS .. _compilers-support-compatibility-matrix-past-60: .. _runtime-support-compatibility-matrix-past-60:
115 :doc:`AMD CLR <hip:understand/amd_clr>` `clang-ocl <https://github.com/ROCm/clang-ocl>`_ N/A 6.3.42131 N/A 6.2.41134 N/A 6.2.41134 N/A 6.2.41134 N/A 6.2.41133 N/A 6.1.40093 0.5.0 6.1.40092 0.5.0 6.1.40091 0.5.0 6.1.32831 0.5.0 6.1.32830 0.5.0
116 :doc:`HIP <hip:index>` :doc:`hipCC <hipcc:index>` 1.1.1 6.3.42131 1.1.1 6.2.41134 1.1.1 6.2.41134 1.1.1 6.2.41134 1.1.1 6.2.41133 1.1.1 6.1.40093 1.0.0 6.1.40092 1.0.0 6.1.40091 1.0.0 6.1.32831 1.0.0 6.1.32830 1.0.0
117 `OpenCL Runtime <https://github.com/ROCm/clr/tree/develop/opencl>`_ `Flang <https://github.com/ROCm/flang>`_ 18.0.0.24491 2.0.0 18.0.0.24455 2.0.0 18.0.0.24392 2.0.0 18.0.0.24355 2.0.0 18.0.0.24355 2.0.0 18.0.0.24232 2.0.0 17.0.0.24193 2.0.0 17.0.0.24154 2.0.0 17.0.0.24103 2.0.0 17.0.0.24012 2.0.0 17.0.0.23483
118 :doc:`ROCr Runtime <rocr-runtime:index>` :doc:`llvm-project <llvm-project:index>` 18.0.0.24455 1.14.0 18.0.0.24491 1.14.0 18.0.0.24392 1.14.0 18.0.0.24355 1.14.0 18.0.0.24355 1.13.0 18.0.0.24232 1.13.0 17.0.0.24193 1.13.0 17.0.0.24154 1.13.0 17.0.0.24103 1.12.0 17.0.0.24012 1.12.0 17.0.0.23483
119 `OpenMP <https://github.com/ROCm/llvm-project/tree/amd-staging/openmp>`_ 18.0.0.24455 18.0.0.24491 18.0.0.24392 18.0.0.24355 18.0.0.24355 18.0.0.24232 17.0.0.24193 17.0.0.24154 17.0.0.24103 17.0.0.24012 17.0.0.23483
120
121
122 RUNTIMES .. _runtime-support-compatibility-matrix-past-60:
123 :doc:`AMD CLR <hip:understand/amd_clr>` 6.3.42133 6.3.42131 6.2.41134 6.2.41134 6.2.41134 6.2.41133 6.1.40093 6.1.40092 6.1.40091 6.1.32831 6.1.32830
124
125 :doc:`HIP <hip:index>` 6.3.42133 6.3.42131 6.2.41134 6.2.41134 6.2.41134 6.2.41133 6.1.40093 6.1.40092 6.1.40091 6.1.32831 6.1.32830
126
127 `OpenCL Runtime <https://github.com/ROCm/clr/tree/develop/opencl>`_ 2.0.0 2.0.0 2.0.0 2.0.0 2.0.0 2.0.0 2.0.0 2.0.0 2.0.0 2.0.0 2.0.0
128 :doc:`ROCr Runtime <rocr-runtime:index>` 1.14.0 1.14.0 1.14.0 1.14.0 1.14.0 1.13.0 1.13.0 1.13.0 1.13.0 1.12.0 1.12.0

View File

@@ -23,17 +23,16 @@ compatibility and system requirements.
.. container:: format-big-table
.. csv-table::
:header: "ROCm Version", "6.3.0", "6.2.4", "6.1.0"
:header: "ROCm Version", "6.3.1", "6.3.0", "6.2.0"
:stub-columns: 1
:ref:`Operating systems & kernels <OS-kernel-versions>`,Ubuntu 24.04.2,"Ubuntu 24.04.1, 24.04",
,Ubuntu 22.04.5,"Ubuntu 22.04.5, 22.04.4","Ubuntu 22.04.4, 22.04.3"
,,,"Ubuntu 20.04.6, 20.04.5"
,"RHEL 9.5, 9.4","RHEL 9.4, 9.3","RHEL 9.4 [#red-hat94]_, 9.3, 9.2"
,"RHEL 8.10","RHEL 8.10, 8.9","RHEL 8.9, 8.8"
,"SLES 15 SP6, SP5","SLES 15 SP6, SP5","SLES 15 SP5, SP4"
,,,CentOS 7.9
,Oracle Linux 8.10 [#oracle89]_,Oracle Linux 8.9 [#oracle89]_,
:ref:`Operating systems & kernels <OS-kernel-versions>`,Ubuntu 24.04.2,Ubuntu 24.04.2,Ubuntu 24.04
,Ubuntu 22.04.5,Ubuntu 22.04.5,"Ubuntu 22.04.5, 22.04.4"
,"RHEL 9.5, 9.4","RHEL 9.5, 9.4","RHEL 9.4, 9.3"
,RHEL 8.10,RHEL 8.10,"RHEL 8.10, 8.9"
,"SLES 15 SP6, SP5","SLES 15 SP6, SP5","SLES 15 SP6, SP5"
,Oracle Linux 8.10 [#mi300x]_,Oracle Linux 8.10 [#mi300x]_,Oracle Linux 8.9 [#mi300x]_
,Debian 12 [#mi300x]_,,
,.. _architecture-support-compatibility-matrix:,,
:doc:`Architecture <rocm-install-on-linux:reference/system-requirements>`,CDNA3,CDNA3,CDNA3
,CDNA2,CDNA2,CDNA2
@@ -43,115 +42,114 @@ compatibility and system requirements.
,.. _gpu-support-compatibility-matrix:,,
:doc:`GPU / LLVM target <rocm-install-on-linux:reference/system-requirements>`,gfx1100,gfx1100,gfx1100
,gfx1030,gfx1030,gfx1030
,gfx942,gfx942 [#mi300_624]_, gfx942 [#mi300_610]_
,gfx942,gfx942,gfx942 [#mi300_620]_
,gfx90a,gfx90a,gfx90a
,gfx908,gfx908,gfx908
,,,
FRAMEWORK SUPPORT,.. _framework-support-compatibility-matrix:,,
:doc:`PyTorch <rocm-install-on-linux:install/3rd-party/pytorch-install>`,"2.4, 2.3, 2.2, 2.1, 2.0, 1.13","2.3, 2.2, 2.1, 2.0, 1.13","2.1, 2.0, 1.13"
:doc:`TensorFlow <rocm-install-on-linux:install/3rd-party/tensorflow-install>`,"2.17.0, 2.16.2, 2.15.1","2.16.1, 2.15.1, 2.14.1","2.15.0, 2.14.0, 2.13.1"
:doc:`JAX <rocm-install-on-linux:install/3rd-party/jax-install>`,0.4.26,0.4.26,0.4.26
:doc:`PyTorch <rocm-install-on-linux:install/3rd-party/pytorch-install>`,"2.4, 2.3, 2.2, 1.13","2.4, 2.3, 2.2, 2.1, 2.0, 1.13","2.3, 2.2, 2.1, 2.0, 1.13"
:doc:`TensorFlow <rocm-install-on-linux:install/3rd-party/tensorflow-install>`,"2.17.0, 2.16.2, 2.15.1","2.17.0, 2.16.2, 2.15.1","2.16.1, 2.15.1, 2.14.1"
:doc:`JAX <rocm-install-on-linux:install/3rd-party/jax-install>`,0.4.35,0.4.35,0.4.26
`ONNX Runtime <https://onnxruntime.ai/docs/build/eps.html#amd-migraphx>`_,1.17.3,1.17.3,1.17.3
,,,
THIRD PARTY COMMS,.. _thirdpartycomms-support-compatibility-matrix:,,
`UCC <https://github.com/ROCm/ucc>`_,>=1.3.0,>=1.3.0,>=1.3.0
`UCX <https://github.com/ROCm/ucx>`_,>=1.15.0,>=1.15.0,>=1.14.1
`UCX <https://github.com/ROCm/ucx>`_,>=1.15.0,>=1.15.0,>=1.15.0
,,,
THIRD PARTY ALGORITHM,.. _thirdpartyalgorithm-support-compatibility-matrix:,,
Thrust,2.3.2,2.2.0,2.1.0
CUB,2.3.2,2.2.0,2.1.0
Thrust,2.3.2,2.3.2,2.2.0
CUB,2.3.2,2.3.2,2.2.0
,,,
KFD & USER SPACE [#kfd_support]_,.. _kfd-userspace-support-compatibility-matrix:,,
Tested user space versions,"6.3.x, 6.2.x, 6.1.x","6.3.x, 6.2.x, 6.1.x, 6.0.x","6.3.x, 6.2.x, 6.1.x, 6.0.x, 5.7.x"
KMD & USER SPACE [#kfd_support]_,.. _kfd-userspace-support-compatibility-matrix:,,
Tested user space versions,"6.3.x, 6.2.x, 6.1.x","6.3.x, 6.2.x, 6.1.x","6.3.x, 6.2.x, 6.1.x, 6.0.x"
,,,
ML & COMPUTER VISION,.. _mllibs-support-compatibility-matrix:,,
:doc:`Composable Kernel <composable_kernel:index>`,1.1.0,1.1.0,1.1.0
:doc:`MIGraphX <amdmigraphx:index>`,2.11.0,2.10.0,2.9.0
:doc:`MIOpen <miopen:index>`,3.3.0,3.2.0,3.1.0
:doc:`MIVisionX <mivisionx:index>`,3.1.0,3.0.0,2.5.0
:doc:`rocAL <rocal:index>`,2.1.0,2.0.0,1.0.0
:doc:`rocDecode <rocdecode:index>`,0.8.0,0.6.0,0.5.0
:doc:`rocJPEG <rocjpeg:index>`,0.6.0,N/A,N/A
:doc:`rocPyDecode <rocpydecode:index>`,0.2.0,0.1.0,N/A
:doc:`RPP <rpp:index>`,1.9.1,1.8.0,1.5.0
:doc:`MIGraphX <amdmigraphx:index>`,2.11.0,2.11.0,2.10.0
:doc:`MIOpen <miopen:index>`,3.3.0,3.3.0,3.2.0
:doc:`MIVisionX <mivisionx:index>`,3.1.0,3.1.0,3.0.0
:doc:`rocAL <rocal:index>`,2.1.0,2.1.0,1.0.0
:doc:`rocDecode <rocdecode:index>`,0.8.0,0.8.0,0.6.0
:doc:`rocJPEG <rocjpeg:index>`,0.6.0,0.6.0,N/A
:doc:`rocPyDecode <rocpydecode:index>`,0.2.0,0.2.0,0.1.0
:doc:`RPP <rpp:index>`,1.9.1,1.9.1,1.8.0
,,,
COMMUNICATION,.. _commlibs-support-compatibility-matrix:,,
:doc:`RCCL <rccl:index>`,2.21.5,2.20.5,2.18.6
:doc:`RCCL <rccl:index>`,2.21.5,2.21.5,2.20.5
,,,
MATH LIBS,.. _mathlibs-support-compatibility-matrix:,,
`half <https://github.com/ROCm/half>`_ ,1.12.0,1.12.0,1.12.0
:doc:`hipBLAS <hipblas:index>`,2.3.0,2.2.0,2.1.0
:doc:`hipBLASLt <hipblaslt:index>`,0.10.0,0.8.0,0.7.0
:doc:`hipFFT <hipfft:index>`,1.0.17,1.0.16,1.0.14
:doc:`hipfort <hipfort:index>`,0.5.0,0.4.0,0.4.0
:doc:`hipRAND <hiprand:index>`,2.11.0,2.11.1,2.10.16
:doc:`hipSOLVER <hipsolver:index>`,2.3.0,2.2.0,2.1.0
:doc:`hipSPARSE <hipsparse:index>`,3.1.2,3.1.1,3.0.1
:doc:`hipSPARSELt <hipsparselt:index>`,0.2.2,0.2.1,0.1.0
:doc:`rocALUTION <rocalution:index>`,3.2.1,3.2.1,3.1.1
:doc:`rocBLAS <rocblas:index>`,4.3.0,4.2.4,4.1.0
:doc:`rocFFT <rocfft:index>`,1.0.31,1.0.30,1.0.26
:doc:`rocRAND <rocrand:index>`,3.2.0,3.1.1,3.0.1
:doc:`rocSOLVER <rocsolver:index>`,3.27.0,3.26.2,3.25.0
:doc:`rocSPARSE <rocsparse:index>`,3.3.0,3.2.1,3.1.2
:doc:`rocWMMA <rocwmma:index>`,1.6.0,1.5.0,1.4.0
:doc:`Tensile <tensile:index>`,4.42.0,4.41.0,4.40.0
:doc:`hipBLAS <hipblas:index>`,2.3.0,2.3.0,2.2.0
:doc:`hipBLASLt <hipblaslt:index>`,0.10.0,0.10.0,0.8.0
:doc:`hipFFT <hipfft:index>`,1.0.17,1.0.17,1.0.14
:doc:`hipfort <hipfort:index>`,0.5.0,0.5.0,0.4.0
:doc:`hipRAND <hiprand:index>`,2.11.1,2.11.0,2.11.0
:doc:`hipSOLVER <hipsolver:index>`,2.3.0,2.3.0,2.2.0
:doc:`hipSPARSE <hipsparse:index>`,3.1.2,3.1.2,3.1.1
:doc:`hipSPARSELt <hipsparselt:index>`,0.2.2,0.2.2,0.2.1
:doc:`rocALUTION <rocalution:index>`,3.2.1,3.2.1,3.2.0
:doc:`rocBLAS <rocblas:index>`,4.3.0,4.3.0,4.2.0
:doc:`rocFFT <rocfft:index>`,1.0.31,1.0.31,1.0.28
:doc:`rocRAND <rocrand:index>`,3.2.0,3.2.0,3.1.0
:doc:`rocSOLVER <rocsolver:index>`,3.27.0,3.27.0,3.26.0
:doc:`rocSPARSE <rocsparse:index>`,3.3.0,3.3.0,3.2.0
:doc:`rocWMMA <rocwmma:index>`,1.6.0,1.6.0,1.5.0
:doc:`Tensile <tensile:index>`,4.42.0,4.42.0,4.41.0
,,,
PRIMITIVES,.. _primitivelibs-support-compatibility-matrix:,,
:doc:`hipCUB <hipcub:index>`,3.3.0,3.2.1,3.1.0
:doc:`hipTensor <hiptensor:index>`,1.4.0,1.3.0,1.2.0
:doc:`rocPRIM <rocprim:index>`,3.3.0,3.2.2,3.1.0
:doc:`rocThrust <rocthrust:index>`,3.3.0,3.1.1,3.0.1
:doc:`hipCUB <hipcub:index>`,3.3.0,3.3.0,3.2.0
:doc:`hipTensor <hiptensor:index>`,1.4.0,1.4.0,1.3.0
:doc:`rocPRIM <rocprim:index>`,3.3.0,3.3.0,3.2.0
:doc:`rocThrust <rocthrust:index>`,3.3.0,3.3.0,3.0.1
,,,
SUPPORT LIBS,,,
`hipother <https://github.com/ROCm/hipother>`_,6.3.42131,6.2.41134,6.1.40091
`rocm-core <https://github.com/ROCm/rocm-core>`_,6.3.0,6.2.4,6.1.0
`ROCT-Thunk-Interface <https://github.com/ROCm/ROCT-Thunk-Interface>`_,N/A [#ROCT-rocr]_,20240607.5.7,20240125.3.30
`hipother <https://github.com/ROCm/hipother>`_,6.3.42133,6.3.42131,6.2.41133
`rocm-core <https://github.com/ROCm/rocm-core>`_,6.3.1,6.3.0,6.2.0
`ROCT-Thunk-Interface <https://github.com/ROCm/ROCT-Thunk-Interface>`_,N/A [#ROCT-rocr]_,N/A [#ROCT-rocr]_,20240607.1.4246
,,,
SYSTEM MGMT TOOLS,.. _tools-support-compatibility-matrix:,,
:doc:`AMD SMI <amdsmi:index>`,24.7.1,24.6.3,24.4.1
:doc:`AMD SMI <amdsmi:index>`,24.7.1,24.7.1,24.6.2
:doc:`ROCm Data Center Tool <rdc:index>`,0.3.0,0.3.0,0.3.0
:doc:`rocminfo <rocminfo:index>`,1.0.0,1.0.0,1.0.0
:doc:`ROCm SMI <rocm_smi_lib:index>`,7.4.0,7.3.0,7.0.0
:doc:`ROCm Validation Suite <rocmvalidationsuite:index>`,1.1.0,1.0.60204,1.0.60100
:doc:`ROCm SMI <rocm_smi_lib:index>`,7.4.0,7.4.0,7.3.0
:doc:`ROCm Validation Suite <rocmvalidationsuite:index>`,1.1.0,1.1.0,1.0.60200
,,,
PERFORMANCE TOOLS,,,
:doc:`ROCm Bandwidth Test <rocm_bandwidth_test:index>`,1.4.0,1.4.0,1.4.0
:doc:`ROCm Compute Profiler <rocprofiler-compute:index>`,3.0.0,2.0.1,N/A
:doc:`ROCm Systems Profiler <rocprofiler-systems:index>`,0.1.0,1.11.2,N/A
:doc:`ROCProfiler <rocprofiler:index>`,2.0.60300,2.0.60204,2.0.60100
:doc:`ROCprofiler-SDK <rocprofiler-sdk:index>`,0.5.0,0.4.0,N/A
:doc:`ROCTracer <roctracer:index>`,4.1.60300,4.1.60204,4.1.60100
:doc:`ROCm Compute Profiler <rocprofiler-compute:index>`,3.0.0,3.0.0,2.0.1
:doc:`ROCm Systems Profiler <rocprofiler-systems:index>`,0.1.0,0.1.0,1.11.2
:doc:`ROCProfiler <rocprofiler:index>`,2.0.60301,2.0.60300,2.0.60200
:doc:`ROCprofiler-SDK <rocprofiler-sdk:index>`,0.5.0,0.5.0,0.4.0
:doc:`ROCTracer <roctracer:index>`,4.1.60301,4.1.60300,4.1.60200
,,,
DEVELOPMENT TOOLS,,,
:doc:`HIPIFY <hipify:index>`,18.0.0.24455,18.0.0.24392,17.0.0.24103
:doc:`ROCm CMake <rocmcmakebuildtools:index>`,0.14.0,0.13.0,0.12.0
:doc:`ROCdbgapi <rocdbgapi:index>`,0.77.0,0.76.0,0.71.0
:doc:`ROCm Debugger (ROCgdb) <rocgdb:index>`,15.2.0,14.2.0,14.1.0
`rocprofiler-register <https://github.com/ROCm/rocprofiler-register>`_,0.4.0,0.4.0,0.3.0
:doc:`HIPIFY <hipify:index>`,18.0.0.24491,18.0.0.24455,18.0.0.24232
:doc:`ROCm CMake <rocmcmakebuildtools:index>`,0.14.0,0.14.0,0.13.0
:doc:`ROCdbgapi <rocdbgapi:index>`,0.77.0,0.77.0,0.76.0
:doc:`ROCm Debugger (ROCgdb) <rocgdb:index>`,15.2.0,15.2.0,14.2.0
`rocprofiler-register <https://github.com/ROCm/rocprofiler-register>`_,0.4.0,0.4.0,0.4.0
:doc:`ROCr Debug Agent <rocr_debug_agent:index>`,2.0.3,2.0.3,2.0.3
,,,
COMPILERS,.. _compilers-support-compatibility-matrix:,,
`clang-ocl <https://github.com/ROCm/clang-ocl>`_,N/A,N/A,0.5.0
:doc:`hipCC <hipcc:index>`,1.1.1,1.1.1,1.0.0
`Flang <https://github.com/ROCm/flang>`_,18.0.0.24455,18.0.0.24392,17.0.0.24103
:doc:`llvm-project <llvm-project:index>`,18.0.0.24455,18.0.0.24392,17.0.0.24103
`OpenMP <https://github.com/ROCm/llvm-project/tree/amd-staging/openmp>`_,18.0.0.24455,18.0.0.24392,17.0.0.24103
`clang-ocl <https://github.com/ROCm/clang-ocl>`_,N/A,N/A,N/A
:doc:`hipCC <hipcc:index>`,1.1.1,1.1.1,1.1.1
`Flang <https://github.com/ROCm/flang>`_,18.0.0.24491,18.0.0.24455,18.0.0.24232
:doc:`llvm-project <llvm-project:index>`,18.0.0.24491,18.0.0.24455,18.0.0.24232
`OpenMP <https://github.com/ROCm/llvm-project/tree/amd-staging/openmp>`_,18.0.0.24491,18.0.0.24455,18.0.0.24232
,,,
RUNTIMES,.. _runtime-support-compatibility-matrix:,,
:doc:`AMD CLR <hip:understand/amd_clr>`,6.3.42131,6.2.41134,6.1.40091
:doc:`HIP <hip:index>`,6.3.42131,6.2.41134,6.1.40091
:doc:`AMD CLR <hip:understand/amd_clr>`,6.3.42133,6.3.42131,6.2.41133
:doc:`HIP <hip:index>`,6.3.42133,6.3.42131,6.2.41133
`OpenCL Runtime <https://github.com/ROCm/clr/tree/develop/opencl>`_,2.0.0,2.0.0,2.0.0
:doc:`ROCr Runtime <rocr-runtime:index>`,1.14.0,1.14.0,1.13.0
.. rubric:: Footnotes
.. [#red-hat94] RHEL 9.4 is supported only on AMD Instinct MI300A.
.. [#oracle89] Oracle Linux is supported only on AMD Instinct MI300X.
.. [#mi300_624] **For ROCm 6.2.4** - MI300X (gfx942) is supported on listed operating systems *except* Ubuntu 22.04.5 [6.8 HWE] and Ubuntu 22.04.4 [6.5 HWE].
.. [#mi300_610] **For ROCm 6.1.0** - MI300A (gfx942) is supported on Ubuntu 22.04.4, RHEL 9.4, RHEL 9.3, RHEL 8.9, and SLES 15 SP5. MI300X (gfx942) is only supported on Ubuntu 22.04.4.
.. [#kfd_support] ROCm provides forward and backward compatibility between the Kernel Fusion Driver (KFD) and its user space software for +/- 2 releases. These are the compatibility combinations that are currently supported.
.. [#ROCT-rocr] As of ROCm 6.3.0, the ROCT Thunk Interface is now included as part of the ROCr runtime package.
.. [#mi300x] Oracle Linux and Debian are supported only on AMD Instinct MI300X.
.. [#mi300_620] **For ROCm 6.2.0** - MI300X (gfx942) is supported on listed operating systems *except* Ubuntu 22.04.5 [6.8 HWE] and Ubuntu 22.04.4 [6.5 HWE].
.. [#kfd_support] ROCm provides forward and backward compatibility between the AMD Kernel-mode GPU Driver (KMD) and its user space software for +/- 2 releases. These are the compatibility combinations that are currently supported.
.. [#ROCT-rocr] Starting from ROCm 6.3.0, the ROCT Thunk Interface is included as part of the ROCr runtime package.
.. _OS-kernel-versions:
@@ -166,35 +164,25 @@ Use this lookup table to confirm which operating system and kernel versions are
:stub-columns: 1
`Ubuntu <https://ubuntu.com/about/release-cycle#ubuntu-kernel-release-cycle>`_, 24.04.2, "6.8 GA, 6.11 HWE"
, 24.04.1, "6.8 GA"
, 24.04, "6.8 GA"
,,
`Ubuntu <https://ubuntu.com/about/release-cycle#ubuntu-kernel-release-cycle>`_, 22.04.5, "5.15 GA, 6.8 HWE"
, 22.04.4, "5.15 GA, 6.5 HWE"
, 22.04.3, "5.15 GA, 6.2 HWE"
, 22.04.2, "5.15 GA, 5.19 HWE"
,,
`Ubuntu <https://ubuntu.com/about/release-cycle#ubuntu-kernel-release-cycle>`_, 20.04.06, "5.15 HWE"
, 20.04.5, "5.15 HWE"
,,
`Red Hat Enterprise Linux (RHEL) <https://access.redhat.com/articles/3078#RHEL9>`_, 9.5, 5.14.0
,9.4, 5.14.0
,9.3, 5.14.0
,9.2, 5.14.0
,,
`Red Hat Enterprise Linux (RHEL) <https://access.redhat.com/articles/3078#RHEL8>`_, 8.10, 4.18.0
,8.9, 4.18.0
,8.8, 4.18.0
,,
`CentOS <https://access.redhat.com/articles/3078#RHEL7>`_, 7.9, 3.10
,,
`SUSE Linux Enterprise Server (SLES) <https://www.suse.com/support/kb/doc/?id=000019587#SLE15SP4>`_, 15 SP6, 6.4.0
,15 SP5, 5.14.21
,15 SP4, 5.14.21
,,
`Oracle Linux <https://blogs.oracle.com/scoter/post/oracle-linux-and-unbreakable-enterprise-kernel-uek-releases>`_, 8.10, 5.15.0
,8.9, 5.15.0
`Azure Linux <https://github.com/microsoft/azurelinux/releases>`_, 3.0, 6.6.60
,,
`Debian <https://www.debian.org/download>`_,12, 6.1
..
Footnotes and ref anchors in below historical tables should be appended with "-past-60", to differentiate from the
@@ -222,7 +210,7 @@ Expand for full historical view of:
.. rubric:: Footnotes
.. [#oracle89-past-60] Oracle Linux is supported only on AMD Instinct MI300X.
.. [#mic300x-past-60] Oracle Linux and Debian are supported only on AMD Instinct MI300X.
.. [#mi300_624-past-60] **For ROCm 6.2.4** - MI300X (gfx942) is supported on listed operating systems *except* Ubuntu 22.04.5 [6.8 HWE] and Ubuntu 22.04.4 [6.5 HWE].
.. [#mi300_622-past-60] **For ROCm 6.2.2** - MI300X (gfx942) is supported on listed operating systems *except* Ubuntu 22.04.5 [6.8 HWE] and Ubuntu 22.04.4 [6.5 HWE].
.. [#mi300_621-past-60] **For ROCm 6.2.1** - MI300X (gfx942) is supported on listed operating systems *except* Ubuntu 22.04.5 [6.8 HWE] and Ubuntu 22.04.4 [6.5 HWE].
@@ -232,5 +220,5 @@ Expand for full historical view of:
.. [#mi300_610-past-60] **For ROCm 6.1.0** - MI300A (gfx942) is supported on Ubuntu 22.04.4, RHEL 9.4, RHEL 9.3, RHEL 8.9, and SLES 15 SP5. MI300X (gfx942) is only supported on Ubuntu 22.04.4.
.. [#mi300_602-past-60] **For ROCm 6.0.2** - MI300A (gfx942) is supported on Ubuntu 22.04.3, RHEL 8.9, and SLES 15 SP5. MI300X (gfx942) is only supported on Ubuntu 22.04.3.
.. [#mi300_600-past-60] **For ROCm 6.0.0** - MI300A (gfx942) is supported on Ubuntu 22.04.3, RHEL 8.9, and SLES 15 SP5. MI300X (gfx942) is only supported on Ubuntu 22.04.3.
.. [#kfd_support-past-60] ROCm provides forward and backward compatibility between the Kernel Fusion Driver (KFD) and its user space software for +/- 2 releases. These are the compatibility combinations that are currently supported.
.. [#ROCT-rocr-past-60] As of ROCm 6.3.0, the ROCT Thunk Interface is now included as part of the ROCr runtime package.
.. [#kfd_support-past-60] ROCm provides forward and backward compatibility between the AMD Kernel-mode GPU Driver (KMD) and its user space software for +/- 2 releases. These are the compatibility combinations that are currently supported.
.. [#ROCT-rocr-past-60] Starting from ROCm 6.3.0, the ROCT Thunk Interface is included as part of the ROCr runtime package.

View File

@@ -1,156 +0,0 @@
.. meta::
:description: How ROCm uses PCIe atomics
:keywords: PCIe, PCIe atomics, atomics, BAR memory, AMD, ROCm
*****************************************************************************
How ROCm uses PCIe atomics
*****************************************************************************
ROCm PCIe feature and overview of BAR memory
================================================================
ROCm is an extension of HSA platform architecture, so it shares the queuing model, memory model,
signaling and synchronization protocols. Platform atomics are integral to perform queuing and
signaling memory operations where there may be multiple-writers across CPU and GPU agents.
The full list of HSA system architecture platform requirements are here:
`HSA Sys Arch Features <http://hsafoundation.com/wp-content/uploads/2021/02/HSA-SysArch-1.2.pdf>`_.
AMD ROCm Software uses the new PCI Express 3.0 (Peripheral Component Interconnect Express [PCIe]
3.0) features for atomic read-modify-write transactions which extends inter-processor synchronization
mechanisms to IO to support the defined set of HSA capabilities needed for queuing and signaling
memory operations.
The new PCIe atomic operations operate as completers for ``CAS`` (Compare and Swap), ``FetchADD``,
``SWAP`` atomics. The atomic operations are initiated by the I/O device which support 32-bit, 64-bit and
128-bit operand which target address have to be naturally aligned to operation sizes.
For ROCm the Platform atomics are used in ROCm in the following ways:
* Update HSA queue's read_dispatch_id: 64 bit atomic add used by the command processor on the
GPU agent to update the packet ID it processed.
* Update HSA queue's write_dispatch_id: 64 bit atomic add used by the CPU and GPU agent to
support multi-writer queue insertions.
* Update HSA Signals -- 64bit atomic ops are used for CPU & GPU synchronization.
The PCIe 3.0 atomic operations feature allows atomic transactions to be requested by, routed through
and completed by PCIe components. Routing and completion does not require software support.
Component support for each is detectable via the Device Capabilities 2 (DevCap2) register. Upstream
bridges need to have atomic operations routing enabled or the atomic operations will fail even though
PCIe endpoint and PCIe I/O devices has the capability to atomic operations.
To do atomic operations routing capability between two or more Root Ports, each associated Root Port
must indicate that capability via the atomic operations routing supported bit in the DevCap2 register.
If your system has a PCIe Express Switch it needs to support atomic operations routing. Atomic
operations requests are permitted only if a component's ``DEVCTL2.ATOMICOP_REQUESTER_ENABLE``
field is set. These requests can only be serviced if the upstream components support atomic operation
completion and/or routing to a component which does. Atomic operations routing support=1, routing
is supported; atomic operations routing support=0, routing is not supported.
An atomic operation is a non-posted transaction supporting 32-bit and 64-bit address formats, there
must be a response for Completion containing the result of the operation. Errors associated with the
operation (uncorrectable error accessing the target location or carrying out the atomic operation) are
signaled to the requester by setting the Completion Status field in the completion descriptor, they are
set to to Completer Abort (CA) or Unsupported Request (UR).
To understand more about how PCIe atomic operations work, see
`PCIe atomics <https://pcisig.com/specifications/pciexpress/specifications/ECN_Atomic_Ops_080417.pdf>`_
`Linux Kernel Patch to pci_enable_atomic_request <https://patchwork.kernel.org/project/linux-pci/patch/1443110390-4080-1-git-send-email-jay@jcornwall.me/>`_
There are also a number of papers which talk about these new capabilities:
* `Atomic Read Modify Write Primitives by Intel <https://www.intel.es/content/dam/doc/white-paper/atomic-read-modify-write-primitives-i-o-devices-paper.pdf>`_
* `PCI express 3 Accelerator White paper by Intel <https://www.intel.sg/content/dam/doc/white-paper/pci-express3-accelerator-white-paper.pdf>`_
* `PCIe Generation 4 Base Specification includes atomic operations <https://astralvx.com/storage/2020/11/PCI_Express_Base_4.0_Rev0.3_February19-2014.pdf>`_
* `Xilinx PCIe Ultrascale White paper <https://docs.xilinx.com/v/u/8OZSA2V1b1LLU2rRCDVGQw>`_
Other I/O devices with PCIe atomics support:
* Mellanox ConnectX-5 InfiniBand Card
* Cray Aries Interconnect
* Xilinx 7 Series Devices
Future bus technology with richer I/O atomics operation Support
* GenZ
New PCIe Endpoints with support beyond AMD Ryzen and EPYC CPU; Intel Haswell or newer CPUs
with PCIe Generation 3.0 support.
* Mellanox Bluefield SOC
* Cavium Thunder X2
In ROCm, we also take advantage of PCIe ID based ordering technology for P2P when the GPU
originates two writes to two different targets:
* Write to another GPU memory
* Write to system memory to indicate transfer complete
They are routed off to different ends of the computer but we want to make sure the write to system
memory to indicate transfer complete occurs AFTER P2P write to GPU has complete.
BAR memory overview
----------------------------------------------------------------------------------------------------
On a Xeon E5 based system in the BIOS we can turn on above 4GB PCIe addressing, if so he need to set
memory-mapped input/output (MMIO) base address (MMIOH base) and range (MMIO high size) in the BIOS.
In the Supermicro system in the system bios you need to see the following
* Advanced->PCIe/PCI/PnP configuration-\> Above 4G Decoding = Enabled
* Advanced->PCIe/PCI/PnP Configuration-\>MMIOH Base = 512G
* Advanced->PCIe/PCI/PnP Configuration-\>MMIO High Size = 256G
When we support Large Bar Capability there is a Large Bar VBIOS which also disable the IO bar.
For GFX9 and Vega10 which have Physical Address up 44 bit and 48 bit Virtual address.
* BAR0-1 registers: 64bit, prefetchable, GPU memory. 8GB or 16GB depending on Vega10 SKU. Must
be placed < 2^44 to support P2P access from other Vega10.
* BAR2-3 registers: 64bit, prefetchable, Doorbell. Must be placed \< 2^44 to support P2P access from
other Vega10.
* BAR4 register: Optional, not a boot device.
* BAR5 register: 32bit, non-prefetchable, MMIO. Must be placed \< 4GB.
Here is how our base address register (BAR) works on GFX 8 GPUs with 40 bit Physical Address Limit ::
11:00.0 Display controller: Advanced Micro Devices, Inc. [AMD/ATI] Fiji [Radeon R9 FURY / NANO
Series] (rev c1)
Subsystem: Advanced Micro Devices, Inc. [AMD/ATI] Device 0b35
Flags: bus master, fast devsel, latency 0, IRQ 119
Memory at bf40000000 (64-bit, prefetchable) [size=256M]
Memory at bf50000000 (64-bit, prefetchable) [size=2M]
I/O ports at 3000 [size=256]
Memory at c7400000 (32-bit, non-prefetchable) [size=256K]
Expansion ROM at c7440000 [disabled] [size=128K]
Legend:
1 : GPU Frame Buffer BAR -- In this example it happens to be 256M, but typically this will be size of the
GPU memory (typically 4GB+). This BAR has to be placed \< 2^40 to allow peer-to-peer access from
other GFX8 AMD GPUs. For GFX9 (Vega GPU) the BAR has to be placed \< 2^44 to allow peer-to-peer
access from other GFX9 AMD GPUs.
2 : Doorbell BAR -- The size of the BAR is typically will be \< 10MB (currently fixed at 2MB) for this
generation GPUs. This BAR has to be placed \< 2^40 to allow peer-to-peer access from other current
generation AMD GPUs.
3 : IO BAR -- This is for legacy VGA and boot device support, but since this the GPUs in this project are
not VGA devices (headless), this is not a concern even if the SBIOS does not setup.
4 : MMIO BAR -- This is required for the AMD Driver SW to access the configuration registers. Since the
reminder of the BAR available is only 1 DWORD (32bit), this is placed \< 4GB. This is fixed at 256KB.
5 : Expansion ROM -- This is required for the AMD Driver SW to access the GPU video-bios. This is
currently fixed at 128KB.
For more information, you can review
`Overview of Changes to PCI Express 3.0 <https://www.mindshare.com/files/resources/PCIe%203-0.pdf>`_.

View File

@@ -1,241 +0,0 @@
<head>
<meta charset="UTF-8">
<meta name="description" content="GPU memory">
<meta name="keywords" content="GPU memory, VRAM, video random access memory, pageable
memory, pinned memory, managed memory, AMD, ROCm">
</head>
# GPU memory
For the HIP reference documentation, see:
* {doc}`hip:doxygen/html/group___memory`
* {doc}`hip:doxygen/html/group___memory_m`
Host memory exists on the host (e.g. CPU) of the machine in random access memory (RAM).
Device memory exists on the device (e.g. GPU) of the machine in video random access memory (VRAM).
Recent architectures use graphics double data rate (GDDR) synchronous dynamic random-access memory (SDRAM)such as GDDR6, or high-bandwidth memory (HBM) such as HBM2e.
## Memory allocation
Memory can be allocated in two ways: pageable memory, and pinned memory.
The following API calls with result in these allocations:
| API | Data location | Allocation |
|--------------------|---------------|------------|
| System allocated | Host | Pageable |
| `hipMallocManaged` | Host | Managed |
| `hipHostMalloc` | Host | Pinned |
| `hipMalloc` | Device | Pinned |
:::{tip}
`hipMalloc` and `hipFree` are blocking calls, however, HIP recently added non-blocking versions `hipMallocAsync` and `hipFreeAsync` which take in a stream as an additional argument.
:::
### Pageable memory
Pageable memory is usually gotten when calling `malloc` or `new` in a C++ application.
It is unique in that it exists on "pages" (blocks of memory), which can be migrated to other memory storage.
For example, migrating memory between CPU sockets on a motherboard, or a system that runs out of space in RAM and starts dumping pages of RAM into the swap partition of your hard drive.
### Pinned memory
Pinned memory (or page-locked memory, or non-pageable memory) is host memory that is mapped into the address space of all GPUs, meaning that the pointer can be used on both host and device.
Accessing host-resident pinned memory in device kernels is generally not recommended for performance, as it can force the data to traverse the host-device interconnect (e.g. PCIe), which is much slower than the on-device bandwidth (>40x on MI200).
Pinned host memory can be allocated with one of two types of coherence support:
:::{note}
In HIP, pinned memory allocations are coherent by default (`hipHostMallocDefault`).
There are additional pinned memory flags (e.g. `hipHostMallocMapped` and `hipHostMallocPortable`).
On MI200 these options do not impact performance.
<!-- TODO: link to programming_manual#memory-allocation-flags -->
For more information, see the section *memory allocation flags* in the HIP Programming Guide: {doc}`hip:how-to/programming_manual`.
:::
Much like how a process can be locked to a CPU core by setting affinity, a pinned memory allocator does this with the memory storage system.
On multi-socket systems it is important to ensure that pinned memory is located on the same socket as the owning process, or else each cache line will be moved through the CPU-CPU interconnect, thereby increasing latency and potentially decreasing bandwidth.
In practice, pinned memory is used to improve transfer times between host and device.
For transfer operations, such as `hipMemcpy` or `hipMemcpyAsync`, using pinned memory instead of pageable memory on host can lead to a ~3x improvement in bandwidth.
:::{tip}
If the application needs to move data back and forth between device and host (separate allocations), use pinned memory on the host side.
:::
### Managed memory
Managed memory refers to universally addressable, or unified memory available on the MI200 series of GPUs.
Much like pinned memory, managed memory shares a pointer between host and device and (by default) supports fine-grained coherence, however, managed memory can also automatically migrate pages between host and device.
The allocation will be managed by AMD GPU driver using the Linux HMM (Heterogeneous Memory Management) mechanism.
If heterogenous memory management (HMM) is not available, then `hipMallocManaged` will default back to using system memory and will act like pinned host memory.
Other managed memory API calls will have undefined behavior.
It is therefore recommended to check for managed memory capability with: `hipDeviceGetAttribute` and `hipDeviceAttributeManagedMemory`.
HIP supports additional calls that work with page migration:
* `hipMemAdvise`
* `hipMemPrefetchAsync`
:::{tip}
If the application needs to use data on both host and device regularly, does not want to deal with separate allocations, and is not worried about maxing out the VRAM on MI200 GPUs (64 GB per GCD), use managed memory.
:::
:::{tip}
If managed memory performance is poor, check to see if managed memory is supported on your system and if page migration (XNACK) is enabled.
:::
## Access behavior
Memory allocations for GPUs behave as follow:
| API | Data location | Host access | Device access |
|--------------------|---------------|--------------|----------------------|
| System allocated | Host | Local access | Unhandled page fault |
| `hipMallocManaged` | Host | Local access | Zero-copy |
| `hipHostMalloc` | Host | Local access | Zero-copy* |
| `hipMalloc` | Device | Zero-copy | Local access |
Zero-copy accesses happen over the Infinity Fabric interconnect or PCI-E lanes on discrete GPUs.
:::{note}
While `hipHostMalloc` allocated memory is accessible by a device, the host pointer must be converted to a device pointer with `hipHostGetDevicePointer`.
Memory allocated through standard system allocators such as `malloc`, can be accessed a device by registering the memory via `hipHostRegister`.
The device pointer to be used in kernels can be retrieved with `hipHostGetDevicePointer`.
Registered memory is treated like `hipHostMalloc` and will have similar performance.
On devices that support and have [](#xnack) enabled, such as the MI250X, `hipHostRegister` is not required as memory accesses are handled via automatic page migration.
:::
### XNACK
Normally, host and device memory are separate and data has to be transferred manually via `hipMemcpy`.
On a subset of GPUs, such as the MI200, there is an option to automatically migrate pages of memory between host and device.
This is important for managed memory, where the locality of the data is important for performance.
Depending on the system, page migration may be disabled by default in which case managed memory will act like pinned host memory and suffer degraded performance.
*XNACK* describes the GPUs ability to retry memory accesses that failed due a page fault (which normally would lead to a memory access error), and instead retrieve the missing page.
This also affects memory allocated by the system as indicated by the following table:
| API | Data location | Host after device access | Device after host access |
|--------------------|---------------|--------------------------|--------------------------|
| System allocated | Host | Migrate page to host | Migrate page to device |
| `hipMallocManaged` | Host | Migrate page to host | Migrate page to device |
| `hipHostMalloc` | Host | Local access | Zero-copy |
| `hipMalloc` | Device | Zero-copy | Local access |
To check if page migration is available on a platform, use `rocminfo`:
```sh
$ rocminfo | grep xnack
Name: amdgcn-amd-amdhsa--gfx90a:sramecc+:xnack-
```
Here, `xnack-` means that XNACK is available but is disabled by default.
Turning on XNACK by setting the environment variable `HSA_XNACK=1` and gives the expected result, `xnack+`:
```sh
$ HSA_XNACK=1 rocminfo | grep xnack
Name: amdgcn-amd-amdhsa--gfx90a:sramecc+:xnack+
```
`hipcc`by default will generate code that runs correctly with both XNACK enabled or disabled.
Setting the `--offload-arch=`-option with `xnack+` or `xnack-` forces code to be only run with XNACK enabled or disabled respectively.
```sh
# Compiled kernels will run regardless if XNACK is enabled or is disabled.
hipcc --offload-arch=gfx90a
# Compiled kernels will only be run if XNACK is enabled with XNACK=1.
hipcc --offload-arch=gfx90a:xnack+
# Compiled kernels will only be run if XNACK is disabled with XNACK=0.
hipcc --offload-arch=gfx90a:xnack-
```
:::{tip}
If you want to make use of page migration, use managed memory. While pageable memory will migrate correctly, it is not a portable solution and can have performance issues if the accessed data isn't page aligned.
:::
### Coherence
* *Coarse-grained coherence* means that memory is only considered up to date at kernel boundaries, which can be enforced through `hipDeviceSynchronize`, `hipStreamSynchronize`, or any blocking operation that acts on the null stream (e.g. `hipMemcpy`).
For example, cacheable memory is a type of coarse-grained memory where an up-to-date copy of the data can be stored elsewhere (e.g. in an L2 cache).
* *Fine-grained coherence* means the coherence is supported while a CPU/GPU kernel is running.
This can be useful if both host and device are operating on the same dataspace using system-scope atomic operations (e.g. updating an error code or flag to a buffer).
Fine-grained memory implies that up-to-date data may be made visible to others regardless of kernel boundaries as discussed above.
| API | Flag | Coherence |
|-------------------------|------------------------------|----------------|
| `hipHostMalloc` | `hipHostMallocDefault` | Fine-grained |
| `hipHostMalloc` | `hipHostMallocNonCoherent` | Coarse-grained |
| API | Flag | Coherence |
|-------------------------|------------------------------|----------------|
| `hipExtMallocWithFlags` | `hipDeviceMallocDefault` | Coarse-grained |
| `hipExtMallocWithFlags` | `hipDeviceMallocFinegrained` | Fine-grained |
| API | `hipMemAdvise` argument | Coherence |
|-------------------------|------------------------------|----------------|
| `hipMallocManaged` | | Fine-grained |
| `hipMallocManaged` | `hipMemAdviseSetCoarseGrain` | Coarse-grained |
| `malloc` | | Fine-grained |
| `malloc` | `hipMemAdviseSetCoarseGrain` | Coarse-grained |
:::{tip}
Try to design your algorithms to avoid host-device memory coherence (e.g. system scope atomics). While it can be a useful feature in very specific cases, it is not supported on all systems, and can negatively impact performance by introducing the host-device interconnect bottleneck.
:::
The availability of fine- and coarse-grained memory pools can be checked with `rocminfo`:
```sh
$ rocminfo
...
*******
Agent 1
*******
Name: AMD EPYC 7742 64-Core Processor
...
Pool Info:
Pool 1
Segment: GLOBAL; FLAGS: FINE GRAINED
...
Pool 3
Segment: GLOBAL; FLAGS: COARSE GRAINED
...
*******
Agent 9
*******
Name: gfx90a
...
Pool Info:
Pool 1
Segment: GLOBAL; FLAGS: COARSE GRAINED
...
```
## System direct memory access
In most cases, the default behavior for HIP in transferring data from a pinned host allocation to device will run at the limit of the interconnect.
However, there are certain cases where the interconnect is not the bottleneck.
The primary way to transfer data onto and off of a GPU, such as the MI200, is to use the onboard System Direct Memory Access engine, which is used to feed blocks of memory to the off-device interconnect (either GPU-CPU or GPU-GPU).
Each GCD has a separate SDMA engine for host-to-device and device-to-host memory transfers.
Importantly, SDMA engines are separate from the computing infrastructure, meaning that memory transfers to and from a device will not impact kernel compute performance, though they do impact memory bandwidth to a limited extent.
The SDMA engines are mainly tuned for PCIe-4.0 x16, which means they are designed to operate at bandwidths up to 32 GB/s.
:::{note}
An important feature of the MI250X platform is the Infinity Fabric™ interconnect between host and device.
The Infinity Fabric interconnect supports improved performance over standard PCIe-4.0 (usually ~50% more bandwidth); however, since the SDMA engine does not run at this speed, it will not max out the bandwidth of the faster interconnect.
:::
The bandwidth limitation can be countered by bypassing the SDMA engine and replacing it with a type of copy kernel known as a "blit" kernel.
Blit kernels will use the compute units on the GPU, thereby consuming compute resources, which may not always be beneficial.
The easiest way to enable blit kernels is to set an environment variable `HSA_ENABLE_SDMA=0`, which will disable the SDMA engine.
On systems where the GPU uses a PCIe interconnect instead of an Infinity Fabric interconnect, blit kernels will not impact bandwidth, but will still consume compute resources.
The use of SDMA vs blit kernels also applies to MPI data transfers and GPU-GPU transfers.

View File

@@ -0,0 +1,57 @@
.. meta::
:description: How ROCm uses PCIe atomics
:keywords: PCIe, PCIe atomics, atomics, Atomic operations, AMD, ROCm
*****************************************************************************
How ROCm uses PCIe atomics
*****************************************************************************
AMD ROCm is an extension of the Heterogeneous System Architecture (HSA). To meet the requirements of an HSA-compliant system, ROCm supports queuing models, memory models, and signaling and synchronization protocols. ROCm can perform atomic Read-Modify-Write (RMW) transactions that extend inter-processor synchronization mechanisms to Input/Output (I/O) devices starting from Peripheral Component Interconnect Express 3.0 (PCIe™ 3.0). It supports the defined HSA capabilities for queuing and signaling memory operations. To learn more about the requirements of an HSA-compliant system, see the
`HSA Platform System Architecture Specification <http://hsafoundation.com/wp-content/uploads/2021/02/HSA-SysArch-1.2.pdf>`_.
ROCm uses platform atomics to perform memory operations like queuing, signaling, and synchronization across multiple CPU, GPU agents, and I/O devices. Platform atomics ensure that atomic operations run synchronously, without interruptions or conflicts, across multiple shared resources.
Platform atomics in ROCm
==============================
Platform atomics enable the set of atomic operations that perform RMW actions across multiple processors, devices, and memory locations so that they run synchronously without interruption. An atomic operation is a sequence of computing instructions run as a single, indivisible unit. These instructions are completed in their entirety without any interruptions. If the instructions can't be completed as a unit without interruption, none of the instructions are run. These operations support 32-bit and 64-bit address formats.
Some of the operations for which ROCm uses platform atomics are:
* Update the HSA queue's ``read_dispatch_id``. The command processor on the GPU agent uses a 64-bit atomic add operation. It updates the packet ID it processed.
* Update the HSA queue's ``write_dispatch_id``. The CPU and GPU agents use a 64-bit atomic add operation. It supports multi-writer queue insertions.
* Update HSA Signals. A 64-bit atomic operation is used for CPU & GPU synchronization.
PCIe for atomic operations
----------------------------
ROCm requires CPUs that support PCIe atomics. Similarly, all connected I/O devices should also support PCIe atomics for optimum compatibility. PCIe supports the ``CAS`` (Compare and Swap), ``FetchADD``, and ``SWAP`` atomic operations across multiple resources. These atomic operations are initiated by the I/O devices that support 32-bit, 64-bit, and 128-bit operands. Likewise, the target memory address where these atomic operations are performed should also be aligned to the size of the operand. This alignment ensures that the operations are performed efficiently and correctly without failure.
When an atomic operation is successful, the requester receives a response of completion along with the operation result. However, any errors associated with the operation are signaled to the requester by updating the Completion Status field. Issues accessing the target location or running the atomic operation are common errors. Depending upon the error, the Completion Status field is updated to Completer Abort (CA) or Unsupported Request (UR). The field is present in the Completion Descriptor.
To learn more about the industry standards and specifications of PCIe, see `PCI-SIG Specification <https://pcisig.com/specifications>`_.
To learn more about PCIe and its capabilities, consult the following white papers:
* `Atomic Read Modify Write Primitives by Intel <https://www.intel.es/content/dam/doc/white-paper/atomic-read-modify-write-primitives-i-o-devices-paper.pdf>`_
* `PCI Express 3 Accelerator White paper by Intel <https://www.intel.sg/content/dam/doc/white-paper/pci-express3-accelerator-white-paper.pdf>`_
* `PCIe Generation 4 Base Specification includes atomic operations <https://astralvx.com/storage/2020/11/PCI_Express_Base_4.0_Rev0.3_February19-2014.pdf>`_
* `Xilinx PCIe Ultrascale White paper <https://docs.xilinx.com/v/u/8OZSA2V1b1LLU2rRCDVGQw>`_
Working with PCIe 3.0 in ROCm
-------------------------------
Starting with PCIe 3.0, atomic operations can be requested, routed through, and completed by PCIe components. Routing and completion do not require software support. Component support for each can be identified by the Device Capabilities 2 (DevCap2) register. Upstream
bridges need to have atomic operations routing enabled. If not enabled, the atomic operations will fail even if the
PCIe endpoint and PCIe I/O devices can perform atomic operations.
If your system uses PCIe switches to connect and enable communication between multiple PCIe components, the switches must also support atomic operations routing.
To enable atomic operations routing between multiple root ports, each root port must support atomic operation routing. This capability can be identified from the atomic operations routing support bit in the DevCap2 register. If the bit has value of 1, routing is supported. Atomic operation requests are permitted only if a component's ``DEVCTL2.ATOMICOP_REQUESTER_ENABLE``
field is set. These requests can only be serviced if the upstream components also support atomic operation completion or if the requests can be routed to a component that supports atomic operation completion.
ROCm uses the PCIe-ID-based ordering technology for peer-to-peer (P2P) data transmission. PCIe-ID-based ordering technology is used when the GPU initiates multiple write operations to different memory locations.
For more information on changes implemented in PCIe 3.0, see `Overview of Changes to PCI Express 3.0 <https://www.mindshare.com/files/resources/PCIe%203-0.pdf>`_.

View File

@@ -30,19 +30,20 @@ if os.environ.get("READTHEDOCS", "") == "True":
project = "ROCm Documentation"
author = "Advanced Micro Devices, Inc."
copyright = "Copyright (c) 2024 Advanced Micro Devices, Inc. All rights reserved."
version = "6.3.0"
release = "6.3.0"
version = "6.3.1"
release = "6.3.1"
setting_all_article_info = True
all_article_info_os = ["linux", "windows"]
all_article_info_author = ""
# pages with specific settings
article_pages = [
{"file": "about/release-notes", "os": ["linux", "windows"], "date": "2024-12-03"},
{"file": "about/release-notes", "os": ["linux", "windows"], "date": "2024-12-20"},
{"file": "how-to/deep-learning-rocm", "os": ["linux"]},
{"file": "how-to/rocm-for-ai/index", "os": ["linux"]},
{"file": "how-to/rocm-for-ai/install", "os": ["linux"]},
{"file": "how-to/rocm-for-ai/train-a-model", "os": ["linux"]},
{"file": "how-to/rocm-for-ai/accelerate-training", "os": ["linux"]},
{"file": "how-to/rocm-for-ai/deploy-your-model", "os": ["linux"]},
{"file": "how-to/rocm-for-ai/hugging-face-models", "os": ["linux"]},
{"file": "how-to/rocm-for-hpc/index", "os": ["linux"]},

150
docs/contribute/building.md Normal file
View File

@@ -0,0 +1,150 @@
<head>
<meta charset="UTF-8">
<meta name="description" content="Building ROCm documentation">
<meta name="keywords" content="documentation, Visual Studio Code, GitHub, command line,
AMD, ROCm">
</head>
# Building documentation
## GitHub
If you open a pull request and scroll down to the summary panel,
there is a commit status section. Next to the line
`docs/readthedocs.com:advanced-micro-devices-demo`, there is a `Details` link.
If you click this, it takes you to the Read the Docs build for your pull request.
![GitHub PR commit status](../data/contribute/commit-status.png)
If you don't see this line, click `Show all checks` to get an itemized view.
## Command line
You can build our documentation via the command line using Python.
See the `build.tools.python` setting in the [Read the Docs configuration file](https://github.com/ROCm/ROCm/blob/develop/.readthedocs.yaml) for the Python version used by Read the Docs to build documentation.
See the [Python requirements file](https://github.com/ROCm/ROCm/blob/develop/docs/sphinx/requirements.txt) for Python packages needed to build the documentation.
Use the Python Virtual Environment (`venv`) and run the following commands from the project root:
```sh
python3 -mvenv .venv
.venv/bin/python -m pip install -r docs/sphinx/requirements.txt
.venv/bin/python -m sphinx -T -E -b html -d _build/doctrees -D language=en docs _build/html
```
Navigate to `_build/html/index.html` and open this file in a web browser.
## Visual Studio Code
With the help of a few extensions, you can create a productive environment to author and test
documentation locally using Visual Studio (VS) Code. Follow these steps to configure VS Code:
1. Install the required extensions:
* Python: `(ms-python.python)`
* Live Server: `(ritwickdey.LiveServer)`
2. Add the following entries to `.vscode/settings.json`.
```json
{
"liveServer.settings.root": "/.vscode/build/html",
"liveServer.settings.wait": 1000,
"python.terminal.activateEnvInCurrentTerminal": true
}
```
* `liveServer.settings.root`: Sets the root of the output website for live previews. Must be changed
alongside the `tasks.json` command.
* `liveServer.settings.wait`: Tells the live server to wait with the update in order to give Sphinx time to
regenerate the site contents and not refresh before the build is complete.
* `python.terminal.activateEnvInCurrentTerminal`: Activates the automatic virtual environment, so you
can build the site from the integrated terminal.
3. Add the following tasks to `.vscode/tasks.json`.
```json
{
"version": "2.0.0",
"tasks": [
{
"label": "Build Docs",
"type": "process",
"windows": {
"command": "${workspaceFolder}/.venv/Scripts/python.exe"
},
"command": "${workspaceFolder}/.venv/bin/python3",
"args": [
"-m",
"sphinx",
"-j",
"auto",
"-T",
"-b",
"html",
"-d",
"${workspaceFolder}/.vscode/build/doctrees",
"-D",
"language=en",
"${workspaceFolder}/docs",
"${workspaceFolder}/.vscode/build/html"
],
"problemMatcher": [
{
"owner": "sphinx",
"fileLocation": "absolute",
"pattern": {
"regexp": "^(?:.*\\.{3}\\s+)?(\\/[^:]*|[a-zA-Z]:\\\\[^:]*):(\\d+):\\s+(WARNING|ERROR):\\s+(.*)$",
"file": 1,
"line": 2,
"severity": 3,
"message": 4
}
},
{
"owner": "sphinx",
"fileLocation": "absolute",
"pattern": {
"regexp": "^(?:.*\\.{3}\\s+)?(\\/[^:]*|[a-zA-Z]:\\\\[^:]*):{1,2}\\s+(WARNING|ERROR):\\s+(.*)$",
"file": 1,
"severity": 2,
"message": 3
}
}
],
"group": {
"kind": "build",
"isDefault": true
}
}
]
}
```
> Implementation detail: two problem matchers were needed to be defined,
> because VS Code doesn't tolerate some problem information being potentially
> absent. While a single regex could match all types of errors, if a capture
> group remains empty (the line number doesn't show up in all warning/error
> messages) but the `pattern` references said empty capture group, VS Code
> discards the message completely.
4. Configure the Python virtual environment (`venv`).
From the Command Palette, run `Python: Create Environment`. Select `venv` environment and
`docs/sphinx/requirements.txt`.
5. Build the docs.
Launch the default build task using one of the following options:
* A hotkey (the default is `Ctrl+Shift+B`)
* Issuing the `Tasks: Run Build Task` from the Command Palette
6. Open the live preview.
Navigate to the site output within VS Code: right-click on `.vscode/build/html/index.html` and
select `Open with Live Server`. The contents should update on every rebuild without having to
refresh the browser.

Binary file not shown.

After

Width:  |  Height:  |  Size: 139 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 30 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 36 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 35 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 242 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 155 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 242 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 173 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 340 KiB

View File

@@ -0,0 +1,99 @@
.. meta::
:description: Learn about BAR configuration in AMD GPUs and ways to troubleshoot physical addressing limit
:keywords: BAR memory, MMIO, GPU memory, Physical Addressing Limit, AMD, ROCm
**************************************
Troubleshoot BAR access limitation
**************************************
Direct Memory Access (DMA) to PCIe devices using Base Address Registers (BARs) can be restricted due to physical addressing limits. These restrictions can result in data access failures between the system components. Peer-to-peer (P2P) DMA is used to access resources such as registers and memory between devices. PCIe devices need memory-mapped input/output (MMIO) space for DMA, and these MMIO spaces are defined in the PCIe BARs.
These BARs are a set of 32-bit or 64-bit registers that are used to define the resources that PCIe devices provide. The CPU and other system devices also use these to access the resources of the PCIe devices. P2P DMA only works when one device can directly access the local BAR memory of another. If the memory address of a BAR memory exceeds the physical addressing limit of a device, the device will not be able to access that BAR. This could be the device's own BAR or the BAR of another device in the system.
If the BAR memory exceeds than the physical addressing limit of the device, the device will not be able to access the remote BAR.
To handle any BAR access issues that might occur, you need to be aware of the physical address limitations of the devices and understand the :ref:`BAR configuration of AMD GPUs <bar-configuration>`. This information is important when setting up additional MMIO apertures for PCIe devices in the system's physical address space.
Handling physical address limitation
=============================================
When a system boots, the system BIOS allocates the physical address space for the components in the system, including system memory and MMIO apertures. On modern 64-bit platforms, there are generally two or more MMIO apertures: one located below 4 GB of physical address space for 32-bit compatibility, and one or more above 4 GB for devices needing more space.
You can control the memory address of the high MMIO aperture from the system BIOS configuration options. This lets you configure the additional MMIO space to align with the physical addressing limit and allows P2P DMA between the devices. For example, if a PCIe device is limited to 44-bit of physical addressing, you should ensure that the MMIO aperture is set below 44-bit in the system physical address space.
There are two ways to handle this:
* Ensure that the high MMIO aperture is within the physical addressing limits of the devices in the system. For example, if the devices have a 44-bit physical addressing limit, set the ``MMIO High Base`` and ``MMIO High size`` options in the BIOS such that the aperture is within the 44-bit address range, and ensure that the ``Above 4G Decoding`` option is Enabled.
* Enable the Input-Output Memory Management Unit (IOMMU). When the IOMMU is enabled in non-passthrough mode, it will create a virtual I/O address space for each device on the system. It also ensures that all virtual addresses created in that space are within the physical addressing limits of the device. For more information on IOMMU, see :doc:`../conceptual/iommu`.
.. _bar-configuration:
BAR configuration for AMD GPUs
================================================
The following table shows how the BARs are configured for AMD GPUs.
.. list-table::
:widths: 25 25 50
:header-rows: 1
* - BAR Type
- Value
- Description
* - BAR0-1 registers
- 64-bit, Prefetchable, GPU memory
- 8 GB or 16 GB depending on GPU. Set to less than 2^44 to support P2P access from other GPUs with a 44-bit physical address limit. Prefetchable memory enables faster read operation for high-performance computing (HPC) by fetching the contiguous data from the same data source even before requested as an anticipation of a future request.
* - BAR2-3 registers
- 64-bit, Prefetchable, Doorbell
- Set to less than 2^44 to support P2P access from other GPUs with a 44-bit physical address limit. As a Doorbell BAR, it indicates to the GPU that a new operation is in its queue to be processed.
* - BAR4 register
- Optional
- Not a boot device
* - BAR5 register
- 32-bit, Non-prefetchable, MMIO
- Is set to less than 4 GB.
Example of BAR usage on AMD GPUs
-------------------------------------
Following is an example configuration of BARs set by the system BIOS on GFX8 GPUs with the 40-bit physical addressing limit:
.. code:: shell
11:00.0 Display controller: Advanced Micro Devices, Inc. [AMD/ATI] Fiji [Radeon R9 FURY / NANO
Series] (rev c1)
Subsystem: Advanced Micro Devices, Inc. [AMD/ATI] Device 0b35
Flags: bus master, fast devsel, latency 0, IRQ 119
Memory at bf40000000 (64-bit, prefetchable) [size=256M]
Memory at bf50000000 (64-bit, prefetchable) [size=2M]
I/O ports at 3000 [size=256]
Memory at c7400000 (32-bit, non-prefetchable) [size=256K]
Expansion ROM at c7440000 [disabled] [size=128K]
Details of the BARs configured in the example are:
**GPU Frame Buffer BAR:** ``Memory at bf40000000 (64-bit, prefetchable) [size=256M]``
The size of the BAR in the example is 256 MB. Generally, it will be the size of the GPU memory (typically 4 GB+). Depending upon the physical address limit and generation of AMD GPUs, the BAR can be set below 2^40, 2^44, or 2^48.
**Doorbell BAR:** ``Memory at bf50000000 (64-bit, prefetchable) [size=2M]``
The size of the BAR should typically be less than 10 MB for this generation of GPUs and has been set to 2 MB in the example. This BAR is placed less than 2^40 to allow peer-to-peer access from other generations of AMD GPUs.
**I/O BAR:** ``I/O ports at 3000 [size=256]``
This is for legacy VGA and boot device support. Because the GPUs used are not connected to a display (VGA devices), this is not a concern, even if it isn't set up in the system BIOS.
**MMIO BAR:** ``Memory at c7400000 (32-bit, non-prefetchable) [size=256K]``
The AMD Driver requires this to access the configuration registers. Since the reminder of the BAR available is only 1 DWORD (32-bit), this is set less than 4 GB. In the example, it is fixed at 256 KB.
**Expansion ROM:** ``Expansion ROM at c7440000 [disabled] [size=128K]``
This is required by the AMD Driver to access the GPU video-BIOS. In the example, it is fixed at 128 KB.

View File

@@ -16,6 +16,8 @@ In this guide, you'll learn about:
- :doc:`Installing ROCm and machine learning frameworks <install>`
- :doc:`Scaling model training <scale-model-training>`
- :doc:`Training a model <train-a-model>`
- :doc:`Running models from Hugging Face <hugging-face-models>`

View File

@@ -0,0 +1,135 @@
.. meta::
:description: How to scale and accelerate model training
:keywords: ROCm, AI, LLM, train, fine-tune, deploy, FSDP, DeepSpeed, LLaMA, tutorial
**********************
Scaling model training
**********************
To train a large-scale model like OpenAI GPT-2 or Meta Llama 2 70B, a single accelerator or GPU cannot store all the
model parameters required for training. This immense scale presents a fundamental challenge: no single GPU or
accelerator can simultaneously store and process the entire model's parameters during training. PyTorch
provides an answer to this computational constraint through its distributed training frameworks.
.. _rocm-for-ai-pytorch-distributed:
PyTorch distributed
===================
Features in ``torch.distributed`` are categorized into three main components:
- `Distributed data-parallel training
<https://pytorch.org/docs/stable/generated/torch.nn.parallel.DistributedDataParallel.html>`_ (DDP)
- `RPC-Based distributed training <https://pytorch.org/docs/stable/rpc.html>`_ (RPC)
- `Collective communication <https://pytorch.org/docs/stable/distributed.html>`_
In this topic, the focus is on the distributed data-parallelism strategy as its the most popular. To get started with DDP,
you need to first understand how to coordinate the model and its training data across multiple accelerators or GPUs.
The DDP workflow on multiple accelerators or GPUs is as follows:
#. Split the current global training batch into small local batches on each GPU. For instance, if you have 8 GPUs and
the global batch is set at 32 samples, each of the 8 GPUs will have a local batch size of 4 samples.
#. Copy the model to every device so each can process its local batches independently.
#. Run a forward pass, then a backward pass, and output the gradient of the weights with respect to the loss of the
model for that local batch. This happens in parallel on multiple devices.
#. Synchronize the local gradients computed by each device and combine them to update the model weights. The updated
weights are then redistributed to each device.
In DDP training, each process or worker owns a replica of the model and processes a batch of data, and then the reducer uses
``allreduce`` to sum up gradients over different workers.
See the following developer blogs for more in-depth explanations and examples.
* `Multi GPU training with DDP — PyTorch Tutorials <https://pytorch.org/tutorials/beginner/ddp_series_multigpu.html>`_
* `Building a decoder transformer model on AMD GPUs — ROCm Blogs
<https://rocm.blogs.amd.com/artificial-intelligence/decoder-transformer/README.html#distributed-training-on-multiple-gpus>`_
.. _rocm-for-ai-pytorch-fsdp:
PyTorch FSDP
------------
As noted in :ref:`PyTorch distributed <rocm-for-ai-pytorch-distributed>`, DDP model weights and optimizer states
are evenly replicated across all workers. Fully Sharded Data Parallel (FSDP) is a type of data parallelism that shards
model parameters, optimizer states, and gradients across DDP ranks.
When training with FSDP, the GPU memory footprint is smaller than when training with DDP across all workers. This makes
training some very large models feasible by allowing larger models or batch sizes to fit on-device. However, this
comes with the cost of increased communication volume. The communication overhead is reduced by internal optimizations
like overlapping communication and computation.
For a high-level overview of how FSDP works, review `Getting started with Fully Sharded Data Parallel
<https://pytorch.org/tutorials/intermediate/FSDP_tutorial.html#how-fsdp-works>`_.
For detailed training steps, see `PyTorch FSDP examples
<https://github.com/pytorch/examples/tree/main/distributed/FSDP>`_.
.. _rocm-for-ai-deepspeed:
DeepSpeed
---------
`DeepSpeed <https://deepspeed.ai>`_ offers system innovations that make large-scale deep learning training effective,
efficient, and easy to use. Innovations such as ZeRO, 3D-Parallelism, DeepSpeed-MoE, ZeRO-Infinity, and so on fall under
the training pillar.
See `Pre-training a large language model with Megatron-DeepSpeed on multiple AMD GPUs
<https://rocm.blogs.amd.com/artificial-intelligence/megatron-deepspeed-pretrain/README.html>`_ for a detailed example of
training with DeepSpeed on an AMD accelerator or GPU.
.. _rocm-for-ai-automatic-mixed-precision:
Automatic mixed precision (AMP)
-------------------------------
As models increase in size, so do the time and memory needed to train them; their cost also increases. Any measure we
can take to reduce training time and memory usage through `automatic mixed precision
<https://pytorch.org/docs/stable/amp.html>`_ (AMP) is highly beneficial for most use cases.
See `Automatic mixed precision in PyTorch using AMD GPUs — ROCm Blogs
<https://rocm.blogs.amd.com/artificial-intelligence/automatic-mixed-precision/README.html#automatic-mixed-precision-in-pytorch-using-amd-gpus>`_
for more information about running AMP on an AMD accelerator.
.. _rocm-for-ai-fine-tune:
Fine-tuning your model
======================
ROCm supports multiple techniques for :ref:`optimizing fine-tuning <fine-tuning-llms-concept-optimizations>`, for
example, LoRA, QLoRA, PEFT, and FSDP.
Learn more about challenges and solutions for model fine-tuning in :doc:`../llm-fine-tuning-optimization/index`.
The following developer blogs showcase examples of fine-tuning a model on an AMD accelerator or GPU.
* Fine-tuning Llama2 with LoRA
* `Fine-tune Llama 2 with LoRA: Customizing a large language model for question-answering
<https://rocm.blogs.amd.com/artificial-intelligence/llama2-lora/README.html>`_
* Fine-tuning Llama2 with QLoRA
* `Enhancing LLM accessibility: A deep dive into QLoRA through fine-tuning Llama 2 on a single AMD GPU
<https://rocm.blogs.amd.com/artificial-intelligence/llama2-Qlora/README.html>`_
* Fine-tuning a BERT-based LLM for a text classification task using JAX
* `LLM distributed supervised fine-tuning with JAX
<https://rocm.blogs.amd.com/artificial-intelligence/distributed-sft-jax/README.html>`_
* Fine-tuning StarCoder using PEFT
* `Instruction fine-tuning of StarCoder with PEFT on multiple AMD GPUs
<https://rocm.blogs.amd.com/artificial-intelligence/starcoder-fine-tune/README.html>`_
* Recipes for fine-tuning Llama2 and 3 with ``llama-recipes``
* `meta-llama/llama-recipes: Scripts for fine-tuning Meta Llama3 with composable FSDP & PEFT methods to cover
single/multi-node GPUs <https://github.com/meta-llama/llama-recipes/tree/main/recipes/quickstart/finetuning>`_

View File

@@ -1,140 +1,503 @@
.. meta::
:description: How to use ROCm for AI
:keywords: ROCm, AI, LLM, train, fine-tune, FSDP, DeepSpeed, LLaMA, tutorial
:description: How to train a model using ROCm Megatron-LM
:keywords: ROCm, AI, LLM, train, Megatron-LM, megatron, Llama, tutorial, docker, torch
****************
Training a model
****************
**************************************
Training a model with ROCm Megatron-LM
**************************************
The following is a brief overview of popular component paths per AI development use-case, such as training, LLMs,
and inferencing.
.. _amd-megatron-lm:
Accelerating model training
===========================
The ROCm Megatron-LM framework is a specialized fork of the robust Megatron-LM, designed to
enable efficient training of large-scale language models on AMD GPUs. By leveraging AMD Instinct™ MI300X
accelerators, AMD Megatron-LM delivers enhanced scalability, performance, and resource utilization for AI
workloads. It is purpose-built to :ref:`support models <amd-megatron-lm-model-support>`
like Meta's Llama 2, Llama 3, and Llama 3.1, enabling developers to train next-generation AI models with greater
efficiency. See the GitHub repository at `<https://github.com/ROCm/Megatron-LM>`__.
To train a large model like GPT2 or Llama 2 70B, a single accelerator or GPU cannot store all the model parameters
required for training. What if you could convert the single-GPU training code to run on multiple accelerators or GPUs?
PyTorch offers distributed training solutions to facilitate this.
For ease of use, AMD provides a ready-to-use Docker image for MI300X accelerators containing essential
components, including PyTorch, PyTorch Lightning, ROCm libraries, and Megatron-LM utilities. It contains the
following software to accelerate training workloads:
.. _rocm-for-ai-pytorch-distributed:
+--------------------------+--------------------------------+
| Software component | Version |
+==========================+================================+
| ROCm | 6.1 |
+--------------------------+--------------------------------+
| PyTorch | 2.4.0 |
+--------------------------+--------------------------------+
| PyTorch Lightning | 2.4.0 |
+--------------------------+--------------------------------+
| Megatron Core | 0.9.0 |
+--------------------------+--------------------------------+
| Transformer Engine | 1.5.0 |
+--------------------------+--------------------------------+
| Flash Attention | v2.6 |
+--------------------------+--------------------------------+
| Transformers | 4.44.0 |
+--------------------------+--------------------------------+
PyTorch distributed
-------------------
Supported features and models
=============================
As of PyTorch 1.6.0, features in ``torch.distributed`` are categorized into three main components:
Megatron-LM provides the following key features to train large language models efficiently:
- `Distributed data-parallel training
<https://pytorch.org/docs/stable/generated/torch.nn.parallel.DistributedDataParallel.html>`_ (DDP)
- Transformer Engine (TE)
- `RPC-Based distributed training <https://pytorch.org/docs/stable/rpc.html>`_ (RPC)
- APEX
- `Collective communication <https://pytorch.org/docs/stable/distributed.html>`_
- GEMM tuning
In this guide, the focus is on the distributed data-parallelism strategy as its the most popular. To get started with DDP,
lets first understand how to coordinate the model and its training data across multiple accelerators or GPUs.
- Torch.compile
The DDP workflow on multiple accelerators or GPUs is as follows:
- 3D parallelism: TP + SP + CP
#. Split the current global training batch into small local batches on each GPU. For instance, if you have 8 GPUs and
the global batch is set at 32 samples, each of the 8 GPUs will have a local batch size of 4 samples.
- Distributed optimizer
#. Copy the model to every device so each device can process its local batches independently.
- Flash Attention (FA) 2
#. Run a forward pass, then a backward pass, and output the gradient of the weights with respect to the loss of the
model for that local batch. This happens in parallel on multiple devices.
- Fused kernels
#. Synchronize the local gradients computed by each device and combine them to update the model weights. The updated
weights are then redistributed to each device.
- Pre-training
In DDP training, each process or worker owns a replica of the model and processes a batch of data, then the reducer uses
``allreduce`` to sum up gradients over different workers.
.. _amd-megatron-lm-model-support:
See the following developer blogs for more in-depth explanations and examples.
The following models are pre-optimized for performance on the AMD Instinct MI300X accelerator.
* `Multi GPU training with DDP — PyTorch Tutorials <https://pytorch.org/tutorials/beginner/ddp_series_multigpu.html>`_
* Llama 2 7B
* `Building a decoder transformer model on AMD GPUs — ROCm Blogs
<https://rocm.blogs.amd.com/artificial-intelligence/decoder-transformer/README.html#distributed-training-on-multiple-gpus>`_
* Llama 2 70B
.. _rocm-for-ai-pytorch-fsdp:
* Llama 3 8B
PyTorch FSDP
------------
* Llama 3 70B
As noted in :ref:`PyTorch distributed <rocm-for-ai-pytorch-distributed>`, in DDP model weights and optimizer states
are evenly replicated across all workers. Fully Sharded Data Parallel (FSDP) is a type of data parallelism that shards
model parameters, optimizer states, and gradients across DDP ranks.
* Llama 3.1 8B
When training with FSDP, the GPU memory footprint is smaller than when training with DDP across all workers. This makes
the training of some very large models feasible by allowing larger models or batch sizes to fit on-device. However, this
comes with the cost of increased communication volume. The communication overhead is reduced by internal optimizations
like overlapping communication and computation.
* Llama 3.1 70B
For a high-level overview of how FSDP works, review `Getting started with Fully Sharded Data Parallel
<https://pytorch.org/tutorials/intermediate/FSDP_tutorial.html#how-fsdp-works>`_.
Prerequisite system validation steps
====================================
For detailed training steps, refer to the `PyTorch FSDP examples
<https://github.com/pytorch/examples/tree/main/distributed/FSDP>`_.
Complete the following system validation and optimization steps to set up your system before starting training.
.. _rocm-for-ai-deepspeed:
Disable NUMA auto-balancing
---------------------------
DeepSpeed
---------
Generally, application performance can benefit from disabling NUMA auto-balancing. However,
it might be detrimental to performance with certain types of workloads.
`DeepSpeed <https://deepspeed.ai>`_ offers system innovations that make large-scale deep learning training effective,
efficient, and easy to use. Innovations such as ZeRO, 3D-Parallelism, DeepSpeed-MoE, ZeRO-Infinity, and so on fall under
the training pillar.
Run the command ``cat /proc/sys/kernel/numa_balancing`` to check your current NUMA (Non-Uniform
Memory Access) settings. Output ``0`` indicates this setting is disabled. If there is no output or
the output is ``1``, run the following command to disable NUMA auto-balancing.
See `Pre-training a large language model with Megatron-DeepSpeed on multiple AMD GPUs — ROCm Blogs
<https://rocm.blogs.amd.com/artificial-intelligence/megatron-deepspeed-pretrain/README.html>`_ for a detailed example of
training with DeepSpeed on an AMD accelerator or GPU.
.. code-block:: shell
.. _rocm-for-ai-automatic-mixed-precision:
sudo sh -c 'echo 0 > /proc/sys/kernel/numa_balancing'
Automatic mixed precision (AMP)
See :ref:`mi300x-disable-numa` for more information.
Hardware verification with ROCm
-------------------------------
As models increase in size, the time and memory needed to train them; that is, their cost also increases. Any measure we
can take to reduce training time and memory usage through `automatic mixed precision
<https://pytorch.org/docs/stable/amp.html>`_ (AMP) is highly beneficial for most use cases.
Use the command ``rocm-smi --setperfdeterminism 1900`` to set the max clock speed up to 1900 MHz
instead of the default 2100 MHz. This can reduce the chance of a PCC event lowering the attainable
GPU clocks. This setting will not be required for new IFWI releases with the production PRC feature.
You can restore this setting to its default value with the ``rocm-smi -r`` command.
See `Automatic mixed precision in PyTorch using AMD GPUs — ROCm Blogs
<https://rocm.blogs.amd.com/artificial-intelligence/automatic-mixed-precision/README.html#automatic-mixed-precision-in-pytorch-using-amd-gpus>`_
for more information about running AMP on an AMD accelerator.
Run the command:
.. _rocm-for-ai-fine-tune:
.. code-block:: shell
Fine-tuning your model
======================
rocm-smi --setperfdeterminism 1900
ROCm supports multiple techniques for :ref:`optimizing fine-tuning <fine-tuning-llms-concept-optimizations>`, for
example, LoRA, QLoRA, PEFT, and FSDP.
See :ref:`mi300x-hardware-verification-with-rocm` for more information.
Learn more about challenges and solutions for model fine-tuning in :doc:`../llm-fine-tuning-optimization/index`.
RCCL Bandwidth Test
-------------------
The following developer blogs showcase examples of how to fine-tune a model on an AMD accelerator or GPU.
ROCm Collective Communications Library (RCCL) is a standalone library of standard collective communication
routines for GPUs. See the :doc:`RCCL documentation <rccl:index>` for more information. Before starting
pre-training, running a RCCL bandwidth test helps ensure that the multi-GPU or multi-node setup is optimized
for efficient distributed training.
* Fine-tuning Llama2 with LoRA
Running the RCCL bandwidth test helps verify that:
* `Fine-tune Llama 2 with LoRA: Customizing a large language model for question-answering — ROCm Blogs
<https://rocm.blogs.amd.com/artificial-intelligence/llama2-lora/README.html>`_
- The GPUs can communicate across nodes or within a single node.
* Fine-tuning Llama2 with QLoRA
- The interconnect (such as InfiniBand, Ethernet, or Infinite fabric) is functioning as expected and
provides adequate bandwidth for communication.
* `Enhancing LLM accessibility: A deep dive into QLoRA through fine-tuning Llama 2 on a single AMD GPU — ROCm Blogs
<https://rocm.blogs.amd.com/artificial-intelligence/llama2-Qlora/README.html>`_
- No hardware setup or cabling issues could affect the communication between GPUs
* Fine-tuning a BERT-based LLM for a text classification task using JAX
Tuning and optimizing hyperparameters
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
* `LLM distributed supervised fine-tuning with JAX — ROCm Blogs
<https://rocm.blogs.amd.com/artificial-intelligence/distributed-sft-jax/README.html>`_
In distributed training, specific hyperparameters related to distributed communication can be tuned based on
the results of the RCCL bandwidth test. These variables are already set in the Docker image:
* Fine-tuning StarCoder using PEFT
.. code-block:: shell
* `Instruction fine-tuning of StarCoder with PEFT on multiple AMD GPUs — ROCm Blogs
<https://rocm.blogs.amd.com/artificial-intelligence/starcoder-fine-tune/README.html>`_
# force all RCCL streams to be high priority
export TORCH_NCCL_HIGH_PRIORITY=1
* Recipes for fine-tuning Llama2 and 3 with ``llama-recipes``
# specify which RDMA interfaces to use for communication
export NCCL_IB_HCA=rdma0,rdma1,rdma2,rdma3,rdma4,rdma5,rdma6,rdma7
# define the Global ID index used in RoCE mode
export NCCL_IB_GID_INDEX=3
# avoid data corruption/mismatch issue that existed in past releases
export RCCL_MSCCL_ENABLE=0
Running the RCCL Bandwidth Test
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
It's recommended you run the RCCL bandwidth test before launching training. It ensures system
performance is sufficient to launch training. RCCL is not included in the AMD Megatron-LM Docker
image; follow the instructions in `<https://github.com/ROCm/rccl-tests>`__ to get started.
See :ref:`mi300x-rccl` for more information.
Run on 8 GPUs (``-g 8``), scanning from 8 bytes to 10 GB:
.. code-block:: shell
./build/all_reduce_perf -b 8 -e 10G -f 2 -g 8
.. image:: ../../data/how-to/rocm-for-ai/rccl-tests-8-gpu.png
:width: 800
Using one MPI process per GPU and ``-g 1`` for performance-oriented runs on both single-node and multi-node is
recommended. So, a run on 8 GPUs looks something like:
.. code-block:: shell
mpirun -np 8 --bind-to numa ./build/all_reduce_perf -b 8 -e 10G -f 2 -g 1
.. image:: ../../data/how-to/rocm-for-ai/rccl-tests-1-mpi-process-per-gpu.png
:width: 800
Running with one MPI process per GPU ensures a one-to-one mapping for CPUs and GPUs, which can be beneficial
for smaller message sizes. This better represents the real-world use of RCCL in deep learning frameworks like
PyTorch and TensorFlow.
Use the following script to run the RCCL test for four MI300X GPU nodes. Modify paths and node addresses as needed.
.. code-block::
/home/$USER/ompi_for_gpu/ompi/bin/mpirun -np 32 -H tw022:8,tw024:8,tw010:8, tw015:8 \
--mca pml ucx \
--mca btl ^openib \
-x NCCL_SOCKET_IFNAME=ens50f0np0 \
-x NCCL_IB_HCA=rdma0:1,rdma1:1,rdma2:1,rdma3:1,rdma4:1,rdma5:1,rdma6:1,rdma7:1 \
-x NCCL_IB_GID_INDEX=3 \
-x NCCL_MIN_NCHANNELS=40 \
-x NCCL_DEBUG=version \
$HOME/rccl-tests/build/all_reduce_perf -b 8 -e 8g -f 2 -g 1
.. image:: ../../data/how-to/rocm-for-ai/rccl-tests-4-mi300x-gpu-nodes.png
:width: 800
.. _mi300x-amd-megatron-lm-training:
Start training on MI300X accelerators
=====================================
The pre-built ROCm Megatron-LM environment allows users to quickly validate system performance, conduct
training benchmarks, and achieve superior performance for models like Llama 2 and Llama 3.1.
Use the following instructions to set up the environment, configure the script to train models, and
reproduce the benchmark results on the MI300X accelerators with the AMD Megatron-LM Docker
image.
.. _amd-megatron-lm-requirements:
Download the Docker image and required packages
-----------------------------------------------
1. Use the following command to pull the Docker image from Docker Hub.
.. code-block:: shell
docker pull rocm/megatron-lm:24.12-dev
2. Launch the Docker container.
.. code-block:: shell
docker run -it --device /dev/dri --device /dev/kfd --network host --ipc host --group-add video --cap-add SYS_PTRACE --security-opt seccomp=unconfined --privileged -v $CACHE_DIR:/root/.cache --name megatron-dev-env rocm/megatron-lm:24.12-dev /bin/bash
3. Clone the ROCm Megatron-LM repository to a local directory and install the required packages on the host machine.
.. code-block:: shell
git clone https://github.com/ROCm/Megatron-LM
cd Megatron-LM
.. note::
This release is validated with ``ROCm/Megatron-LM`` commit `bb93ccb <https://github.com/ROCm/Megatron-LM/tree/bb93ccbfeae6363c67b361a97a27c74ab86e7e92>`_.
Checking out this specific commit is recommended for a stable and reproducible environment.
.. code-block:: shell
git checkout bb93ccbfeae6363c67b361a97a27c74ab86e7e92
Prepare training datasets
-------------------------
If you already have the preprocessed data, you can skip this section.
Use the following command to process datasets. We use GPT data as an example. You may change the merge table, use an
end-of-document token, remove sentence splitting, and use the tokenizer type.
.. code-block:: shell
python tools/preprocess_data.py \
--input my-corpus.json \
--output-prefix my-gpt2 \
--vocab-file gpt2-vocab.json \
--tokenizer-type GPT2BPETokenizer \
--merge-file gpt2-merges.txt \
--append-eod
In this case, the automatically generated output files are named ``my-gpt2_text_document.bin`` and
``my-gpt2_text_document.idx``.
.. image:: ../../data/how-to/rocm-for-ai/prep-training-datasets-my-gpt2-text-document.png
:width: 800
.. _amd-megatron-lm-environment-setup:
Environment setup
-----------------
In the ``examples/llama`` directory of Megatron-LM, if you're working with Llama 2 7B or Llama 2 70 B, use the
``train_llama2.sh`` configuration script. Likewise, if you're working with Llama 3 or Llama 3.1, then use
``train_llama3.sh`` and update the configuration script accordingly.
Network interface
^^^^^^^^^^^^^^^^^
To avoid connectivity issues, ensure the correct network interface is set in your training scripts.
1. Run the following command to find the active network interface on your system.
.. code-block:: shell
ip a
2. Update the ``NCCL_SOCKET_IFNAME`` and ``GLOO_SOCKET_IFNAME`` variables with your systems network interface. For
example:
.. code-block:: shell
export NCCL_SOCKET_IFNAME=ens50f0np0
export GLOO_SOCKET_IFNAME=ens50f0np0
Dataset options
^^^^^^^^^^^^^^^
You can use either mock data or real data for training.
* If you're using a real dataset, update the ``DATA_PATH`` variable to point to the location of your dataset.
.. code-block:: shell
DATA_DIR="/root/.cache/data" # Change to where your dataset is stored
DATA_PATH=${DATA_DIR}/bookcorpus_text_sentence
.. code-block:: shell
--data-path $DATA_PATH
Ensure that the files are accessible inside the Docker container.
* Mock data can be useful for testing and validation. If you're using mock data, replace ``--data-path $DATA_PATH`` with the ``--mock-data`` option.
.. code-block:: shell
--mock-data
Tokenizer
^^^^^^^^^
Tokenization is the process of converting raw text into tokens that can be processed by the model. For Llama
models, this typically involves sub-word tokenization, where words are broken down into smaller units based on
a fixed vocabulary. The tokenizer is trained along with the model on a large corpus of text, and it learns a
fixed vocabulary that can represent a wide range of text from different domains. This allows Llama models to
handle a variety of input sequences, including unseen words or domain-specific terms.
To train any of the Llama 2 models that this Docker image supports, use the ``Llama2Tokenizer``.
To train any of Llama 3 and Llama 3.1 models that this Docker image supports, use the ``HuggingFaceTokenizer``.
Set the Hugging Face model link in the ``TOKENIZER_MODEL`` variable.
For example, if you're using the Llama 3.1 8B model:
.. code-block:: shell
TOKENIZER_MODEL=meta-llama/Llama-3.1-8B
Run benchmark tests
-------------------
.. note::
If you're running **multi node training**, update the following environment variables. They can
also be passed as command line arguments.
* Change ``localhost`` to the master node's hostname:
.. code-block:: shell
MASTER_ADDR="${MASTER_ADDR:-localhost}"
* Set the number of nodes you want to train on (for instance, ``2``, ``4``, ``8``):
.. code-block:: shell
NNODES="${NNODES:-1}"
* Set the rank of each node (0 for master, 1 for the first worker node, and so on):
.. code-block:: shell
NODE_RANK="${NODE_RANK:-0}"
* Use this command to run a performance benchmark test of any of the Llama 2 models that this Docker image supports (see :ref:`variables <amd-megatron-lm-benchmark-test-vars>`).
.. code-block:: shell
{variables} bash examples/llama/train_llama2.sh
* Use this command to run a performance benchmark test of any of the Llama 3 and Llama 3.1 models that this Docker image supports (see :ref:`variables <amd-megatron-lm-benchmark-test-vars>`).
.. code-block:: shell
{variables} bash examples/llama/train_llama3.sh
.. _amd-megatron-lm-benchmark-test-vars:
The benchmark tests support the same set of variables:
+--------------------------+-----------------------+-----------------------+
| Name | Options | Description |
+==========================+=======================+=======================+
| ``TEE_OUTPUT`` | 0 or 1 | 0: disable training |
| | | log |
| | | |
| | | 1: enable training |
| | | log |
+--------------------------+-----------------------+-----------------------+
| ``MBS`` | | Micro batch size |
+--------------------------+-----------------------+-----------------------+
| ``BS`` | | Batch size |
+--------------------------+-----------------------+-----------------------+
| ``TP`` | 1, 2, 4, 8 | Tensor parallel |
+--------------------------+-----------------------+-----------------------+
| ``TE_FP8`` | 0 or 1 | Datatype. |
| | | If it is set to 1, |
| | | FP8. |
| | | |
| | | If it is set to 0. |
| | | BP16 |
+--------------------------+-----------------------+-----------------------+
| ``NO_TORCH_COMPILE`` | 0 or 1 | If it is set to 1, |
| | | enable torch.compile. |
| | | |
| | | If it is set to 0. |
| | | Disable torch.compile |
| | | (default) |
+--------------------------+-----------------------+-----------------------+
| ``SEQ_LENGTH`` | | Input sequence length |
+--------------------------+-----------------------+-----------------------+
| ``GEMM_TUNING`` | 0 or 1 | If it is set to 1, |
| | | enable gemm tuning. |
| | | |
| | | If it is set to 0, |
| | | disable gemm tuning |
+--------------------------+-----------------------+-----------------------+
| ``USE_FLASH_ATTN`` | 0 or 1 | 0: disable flash |
| | | attention |
| | | |
| | | 1: enable flash |
| | | attention |
+--------------------------+-----------------------+-----------------------+
| ``ENABLE_PROFILING`` | 0 or 1 | 0: disable torch |
| | | profiling |
| | | |
| | | 1: enable torch |
| | | profiling |
+--------------------------+-----------------------+-----------------------+
| ``MODEL_SIZE`` | | The size of the mode: |
| | | 7B/70B, etc. |
+--------------------------+-----------------------+-----------------------+
| ``TOTAL_ITERS`` | | Total number of |
| | | iterations |
+--------------------------+-----------------------+-----------------------+
| ``transformer-impl`` | transformer_engine or | Enable transformer |
| | local | engine by default |
+--------------------------+-----------------------+-----------------------+
Benchmarking examples
^^^^^^^^^^^^^^^^^^^^^
.. tab-set::
.. tab-item:: Single node training
:sync: single
Use this command to run training with Llama 2 7B model on a single node. You can specify MBS, BS, FP,
datatype, and so on.
.. code-block:: bash
TEE_OUTPUT=1 MBS=5 BS=120 TP=8 TE_FP8=0 NO_TORCH_COMPILE=1
SEQ_LENGTH=4096 bash examples/llama/train_llama2.sh
You can find the training logs at the location defined in ``$TRAIN_LOG`` in the :ref:`configuration script <amd-megatron-lm-environment-setup>`.
See the sample output:
.. image:: ../../data/how-to/rocm-for-ai/llama2-7b-training-log-sample.png
:width: 800
.. tab-item:: Multi node training
:sync: multi
Launch the Docker container on each node.
In this example, run training with Llama 2 7B model on 2 nodes with specific MBS, BS, FP, datatype, and
so on.
On the master node:
.. code-block:: bash
TEE_OUTPUT=1 MBS=4 BS=64 TP=8 TE_FP8=0 NO_TORCH_COMPILE=1
SEQ_LENGTH=4096 bash examples/llama/train_llama2.sh
On the worker node:
.. code-block:: bash
TEE_OUTPUT=1 MBS=4 BS=64 TP=8 TE_FP8=0 NO_TORCH_COMPILE=1
SEQ_LENGTH=4096 bash examples/llama/train_llama2.sh
You can find the training logs at the location defined in ``$TRAIN_LOG`` in the :ref:`configuration script <amd-megatron-lm-environment-setup>`.
Sample output for 2-node training:
Master node:
.. image:: ../../data/how-to/rocm-for-ai/2-node-training-master.png
:width: 800
Worker node:
.. image:: ../../data/how-to/rocm-for-ai/2-node-training-worker.png
:width: 800
* `meta-llama/llama-recipes: Scripts for fine-tuning Meta Llama3 with composable FSDP & PEFT methods to cover
single/multi-node GPUs <https://github.com/meta-llama/llama-recipes/tree/main/recipes/quickstart/finetuning>`_

View File

@@ -537,6 +537,8 @@ installation was successful, refer to the
:doc:`rocm-install-on-linux:install/post-install`.
Should verification fail, consult :doc:`/how-to/system-debugging`.
.. _mi300x-hardware-verification-with-rocm:
Hardware verification with ROCm
-------------------------------

View File

@@ -2062,11 +2062,10 @@ collectives.
Multi-node FSDP and RCCL settings
---------------------------------
It's recommended to use high-priority HIP streams with RCCL.
The simplest way to enable this is by using the nightly PyTorch wheels, as the required changes from
`PR #122830 <https://github.com/pytorch/pytorch/pull/122830>`_ were not included in the PyTorch 2.3
release but are available in the nightly builds.
When using PyTorch's FSDP (Full Sharded Data Parallel) feature, the HIP
streams used by RCCL and HIP streams used for compute kernels do not
always overlap well. As a workaround, it's recommended to use
high-priority HIP streams with RCCL.
To configure high-priority streams:

View File

@@ -37,7 +37,6 @@ ROCm documentation is organized into the following categories:
:::{grid-item-card} How to
:class-body: rocm-card-banner rocm-hue-12
* [Programming guide](./how-to/hip_programming_guide.rst)
* [Use ROCm for AI](./how-to/rocm-for-ai/index.rst)
* [Use ROCm for HPC](./how-to/rocm-for-hpc/index.rst)
* [Fine-tune LLMs and inference optimization](./how-to/llm-fine-tuning-optimization/index.rst)
@@ -47,7 +46,8 @@ ROCm documentation is organized into the following categories:
* [System debugging](./how-to/system-debugging.md)
* [Use MPI](./how-to/gpu-enabled-mpi.rst)
* [Use advanced compiler features](./conceptual/compiler-topics.md)
* [Set the number of CUs](./how-to/setting-cus)
* [Set the number of CUs](./how-to/setting-cus)
* [Troubleshoot BAR access limitation](./how-to/Bar-Memory.rst)
* [ROCm examples](https://github.com/amd/rocm-examples)
:::
@@ -55,12 +55,11 @@ ROCm documentation is organized into the following categories:
:class-body: rocm-card-banner rocm-hue-8
* [GPU architecture overview](./conceptual/gpu-arch.md)
* [GPU memory](./conceptual/gpu-memory.md)
* [Input-Output Memory Management Unit (IOMMU)](./conceptual/iommu.rst)
* [File structure (Linux FHS)](./conceptual/file-reorg.md)
* [GPU isolation techniques](./conceptual/gpu-isolation.md)
* [Using CMake](./conceptual/cmake-packages.rst)
* [ROCm & PCIe atomics](./conceptual/More-about-how-ROCm-uses-PCIe-Atomics.rst)
* [PCIe atomics in ROCm](./conceptual/pcie-atomics.rst)
* [Inception v3 with PyTorch](./conceptual/ai-pytorch-inception.md)
* [Oversubscription of hardware resources](./conceptual/oversubscription.rst)
:::
@@ -73,6 +72,7 @@ ROCm documentation is organized into the following categories:
* [ROCm tools, compilers, and runtimes](./reference/rocm-tools.md)
* [Accelerator and GPU hardware specifications](./reference/gpu-arch-specs.rst)
* [Precision support](./reference/precision-support.rst)
* [Graph safe support](./reference/graph-safe-support.rst)
:::
<!-- markdownlint-enable MD051 -->

View File

@@ -32,6 +32,21 @@ For more information about ROCm hardware compatibility, see the ROCm `Compatibil
- L1 Instruction Cache (KiB)
- VGPR File (KiB)
- SGPR File (KiB)
*
- MI325X
- CDNA3
- gfx942
- 256
- 304 (38 per XCD)
- 64
- 64
- 256
- 32 (4 per XCD)
- 32
- 16 per 2 CUs
- 64 per 2 CUs
- 512
- 12.5
*
- MI300X
- CDNA3

View File

@@ -0,0 +1,111 @@
.. meta::
:description: This page lists supported graph safe ROCm libraries.
:keywords: AMD, ROCm, HIP, hipGRAPH
********************************************************************************
Graph-safe support for ROCm libraries
********************************************************************************
HIP graph-safe libraries operate safely in HIP execution graphs.
:ref:`hip:how_to_HIP_graph` are an alternative way of executing tasks on a GPU
that can provide performance benefits over launching kernels using the standard
method via streams.
Functions and routines from graph-safe libraries shouldnt result in issues like
race conditions, deadlocks, or unintended dependencies.
The following table shows whether a ROCm library is graph-safe.
.. list-table::
:header-rows: 1
*
- ROCm library
- Graph safe support
*
- `Composable Kernel <https://github.com/ROCm/composable_kernel>`_
-
*
- `hipBLAS <https://github.com/ROCm/hipBLAS>`_
-
*
- `hipBLASLt <https://github.com/ROCm/hipBLASLt>`_
- ⚠️
*
- `hipCUB <https://github.com/ROCm/hipCUB>`_
-
*
- `hipFFT <https://github.com/ROCm/hipFFT>`_
- ✅ (see :ref:`details <hipfft:hip-graph-support-for-hipfft>`)
*
- `hipRAND <https://github.com/ROCm/hipRAND>`_
-
*
- `hipSOLVER <https://github.com/ROCm/hipSOLVER>`_
- ⚠️ (experimental)
*
- `hipSPARSE <https://github.com/ROCm/hipSPARSE>`_
-
*
- `hipSPARSELt <https://github.com/ROCm/hipSPARSELt>`_
- ⚠️ (experimental)
*
- `hipTensor <https://github.com/ROCm/hipTensor>`_
-
*
- `MIOpen <https://github.com/ROCm/MIOpen>`_
-
*
- `RCCL <https://github.com/ROCm/rccl>`_
-
*
- `rocAL <https://github.com/ROCm/rocAL>`_
-
*
- `rocALUTION <https://github.com/ROCm/rocALUTION>`_
-
*
- `rocBLAS <https://github.com/ROCm/rocBLAS>`_
- ✅ (see :doc:`details <rocblas:reference/beta-features>`)
*
- `rocDecode <https://github.com/ROCm/rocDecode>`_
-
*
- `rocFFT <https://github.com/ROCm/rocFFT>`_
- ✅ (see :ref:`details <rocfft:hip-graph-support-for-rocfft>`)
*
- `rocHPCG <https://github.com/ROCm/rocHPCG>`_
-
*
- `rocJPEG <https://github.com/ROCm/rocJPEG>`_
-
*
- `rocPRIM <https://github.com/ROCm/rocPRIM>`_
-
*
- `rocRAND <https://github.com/ROCm/rocRAND>`_
-
*
- `rocSOLVER <https://github.com/ROCm/rocSOLVER>`_
- ⚠️ (experimental)
*
- `rocSPARSE <https://github.com/ROCm/rocSPARSE>`_
- ⚠️ (experimental)
*
- `rocThrust <https://github.com/ROCm/rocThrust>`_
- ❌ (see :doc:`details <rocthrust:hipgraph-support>`)
*
- `rocWMMA <https://github.com/ROCm/rocWMMA>`_
-
*
- `RPP <https://github.com/ROCm/rpp>`_
- ⚠️
*
- `Tensile <https://github.com/ROCm/Tensile>`_
-
✅: full support
⚠️: partial support
❌: not supported

View File

@@ -1,7 +1,7 @@
<head>
<meta charset="UTF-8">
<meta name="description" content="ROCm API libraries & tools">
<meta name="keywords" content="ROCm, API, libraries, tools, artificial intelligence, development,
<meta name="keywords" content="ROCm, API, libraries, tools, AI, artificial intelligence, development,
Communications, C++ primitives, Fast Fourier transforms, FFTs, random number generators, linear
algebra, AMD">
</head>

View File

@@ -8,6 +8,7 @@
| Version | Release date |
| ------- | ------------ |
| [6.3.1](https://rocm.docs.amd.com/en/docs-6.3.1/) | December 20, 2024 |
| [6.3.0](https://rocm.docs.amd.com/en/docs-6.3.0/) | December 3, 2024 |
| [6.2.4](https://rocm.docs.amd.com/en/docs-6.2.4/) | November 6, 2024 |
| [6.2.2](https://rocm.docs.amd.com/en/docs-6.2.2/) | September 27, 2024 |

View File

@@ -40,6 +40,8 @@ subtrees:
title: Installation
- file: how-to/rocm-for-ai/train-a-model.rst
title: Train a model
- file: how-to/rocm-for-ai/scale-model-training.rst
title: Scale model training
- file: how-to/rocm-for-ai/hugging-face-models.rst
title: Run models from Hugging Face
- file: how-to/rocm-for-ai/deploy-your-model.rst
@@ -108,7 +110,9 @@ subtrees:
- url: https://rocm.docs.amd.com/projects/llvm-project/en/latest/conceptual/openmp.html
title: OpenMP support
- file: how-to/setting-cus
title: Set the number of CUs
title: Set the number of CUs
- file: how-to/Bar-Memory.rst
title: Troubleshoot BAR access limitation
- url: https://github.com/amd/rocm-examples
title: ROCm examples
@@ -144,8 +148,6 @@ subtrees:
title: AMD Instinct MI100/CDNA1 ISA
- url: https://www.amd.com/system/files/documents/amd-cdna-whitepaper.pdf
title: White paper
- file: conceptual/gpu-memory.md
title: GPU memory
- file: conceptual/iommu.rst
title: Input-Output Memory Management Unit (IOMMU)
- file: conceptual/file-reorg.md
@@ -154,8 +156,8 @@ subtrees:
title: GPU isolation techniques
- file: conceptual/cmake-packages.rst
title: Using CMake
- file: conceptual/More-about-how-ROCm-uses-PCIe-Atomics.rst
title: ROCm & PCIe atomics
- file: conceptual/pcie-atomics.rst
title: PCIe atomics in ROCm
- file: conceptual/ai-pytorch-inception.md
title: Inception v3 with PyTorch
- file: conceptual/oversubscription.rst
@@ -171,6 +173,8 @@ subtrees:
title: Hardware specifications
- file: reference/precision-support.rst
title: Precision support
- file: reference/graph-safe-support.rst
title: Graph safe support
- caption: Contribute
entries:

View File

@@ -10,9 +10,9 @@ ROCm is a software stack, composed primarily of open-source software, that
provides the tools for programming AMD Graphics Processing Units (GPUs), from
low-level kernels to high-level end-user applications.
.. image:: data/rocm-software-stack-6_3_0.jpg
.. image:: data/rocm-software-stack-6_3_1.jpg
:width: 800
:alt: AMD's ROCm software stack and neighboring technologies.
:alt: AMD's ROCm software stack and enabling technologies.
:align: center
Specifically, ROCm provides the tools for

View File

@@ -1,7 +1,7 @@
<?xml version="1.0" encoding="UTF-8"?>
<manifest>
<remote name="rocm-org" fetch="https://github.com/ROCm/" />
<default revision="refs/tags/rocm-6.3.0"
<default revision="refs/tags/rocm-6.3.1"
remote="rocm-org"
sync-c="true"
sync-j="4" />

View File

@@ -0,0 +1,61 @@
# ROCm 6.3.1 release notes
The release notes provide a summary of notable changes since the previous ROCm release.
- [Release highlights](#release-highlights)
- [Operating system and hardware support changes](#operating-system-and-hardware-support-changes)
- [ROCm components versioning](#rocm-components)
- [Detailed component changes](#detailed-component-changes)
- [ROCm known issues](#rocm-known-issues)
- [ROCm resolved issues](#rocm-resolved-issues)
- [ROCm upcoming changes](#rocm-upcoming-changes)
```{note}
If youre using Radeon™ PRO or Radeon GPUs in a workstation setting with a
display connected, continue to use ROCm 6.2.3. See the [Use ROCm on Radeon GPUs](https://rocm.docs.amd.com/projects/radeon/en/latest/index.html)
documentation to verify compatibility and system requirements.
```
## Release highlights
The following are notable new features and improvements in ROCm 6.3.1. For changes to individual components, see
[Detailed component changes](#detailed-component-changes).
### Per queue resiliency for Instinct MI300 accelerators
The AMDGPU driver now includes enhanced resiliency for misbehaving applications on AMD Instinct MI300 accelerators. This helps isolate the impact of misbehaving applications, ensuring other workloads running on the same accelerator are unaffected.
### ROCm Runfile Installer
ROCm 6.3.1 introduces the ROCm Runfile Installer, with initial support for Ubuntu 22.04. The ROCm Runfile Installer facilitates ROCm installation without using a native Linux package management system, with or without network or internet access. For more information, see the [ROCm Runfile Installer documentation](https://rocm.docs.amd.com/projects/install-on-linux/en/docs-6.3.1/install/rocm-runfile-installer.html).
### ROCm documentation updates
ROCm documentation continues to be updated to provide clearer and more comprehensive guidance for a wider variety of user needs and use cases.
* Added documentation on training a model with ROCm Megatron-LM. AMD offers a Docker image for MI300X accelerators
containing essential components to get started, including ROCm libraries, PyTorch, and Megatron-LM utilities. See
[Training a model using ROCm Megatron-LM](https://rocm.docs.amd.com/en/latest/how-to/rocm-for-ai/train-a-model.html)
to get started.
The new ROCm Megatron-LM training Docker accompanies the [ROCm vLLM inference
Docker](https://rocm.docs.amd.com/en/latest/how-to/performance-validation/mi300x/vllm-benchmark.html)
as a set of ready-to-use containerized solutions to get started with using ROCm
for AI.
* Updated the [Instinct MI300X workload tuning
guide](https://rocm.docs.amd.com/en/latest/how-to/tuning-guides/mi300x/workload.html) with more current optimization
strategies. The updated sections include guidance on vLLM optimization, PyTorch TunableOp, and hipBLASLt tuning.
* HIP graph-safe libraries operate safely in HIP execution graphs. [HIP graphs](https://rocm.docs.amd.com/projects/HIP/en/latest/how-to/hip_runtime_api/hipgraph.html#how-to-hip-graph) are an alternative way of executing tasks on a GPU that can provide performance benefits over launching kernels using the standard method via streams. A topic that shows whether a [ROCm library is graph-safe](https://advanced-micro-devices-demo--3953.com.readthedocs.build/en/3953/reference/graph-safe-support.html) has been added.
* The [Device memory](https://rocm.docs.amd.com/projects/HIP/en/latest/how-to/hip_runtime_api/memory_management/device_memory.html) topic in the HIP memory management section has been updated.
* The HIP documentation has expanded with new resources for developers:
* [Multi device management](https://rocm.docs.amd.com/projects/HIP/en/latest/how-to/hip_runtime_api/multi_device.html)
* [OpenGL interoperability](https://rocm.docs.amd.com/projects/HIP/en/latest/how-to/hip_runtime_api/opengl_interop.html)

View File

@@ -0,0 +1,8 @@
## ROCm known issues
ROCm known issues are noted on {fab}`github` [GitHub](https://github.com/ROCm/ROCm/labels/Verified%20Issue). For known
issues related to individual components, review the [Detailed component changes](#detailed-component-changes).
### PCI Express Qualification Tool failure on Debian 12
The PCI Express Qualification Tool (PEQT) module present in the ROCm Validation Suite (RVS) might fail due to the segmentation issue in Debian 12 (bookworm). This will result in failure to determine the characteristics of the PCIe interconnect between the host platform and the GPU like support for Gen 3 atomic completers, DMA transfer statistics, link speed, and link width. The standard PCIe command `lspci` can be used as an alternative to view the characteristics of the PCIe bus interconnect with the GPU. This issue is under investigation and will be addressed in a future release. See [GitHub issue #4175](https://github.com/ROCm/ROCm/issues/4175).

View File

@@ -0,0 +1,23 @@
## ROCm resolved issues
The following are previously known issues resolved in this release. For resolved issues related to
individual components, review the [Detailed component changes](#detailed-component-changes).
### Instinct MI300 series: backward weights convolution performance issue
Fixed a performance issue affecting certain tensor shapes during backward weights convolution when using FP16 or FP32 data types on Instinct MI300 series accelerators. See [GitHub issue #4080](https://github.com/ROCm/ROCm/issues/4080).
### ROCm Compute Profiler and ROCm Systems Profiler post-upgrade issues
Packaging metadata for ROCm Compute Profiler (`rocprofiler-compute`) and ROCm Systems Profiler
(`rocprofiler-systems`) has been updated to handle the renaming from Omniperf and Omnitrace,
respectively. This fixes minor issues when upgrading from ROCm 6.2 to 6.3. For more information, see the GitHub issues
[#4082](https://github.com/ROCm/ROCm/issues/4082) and
[#4083](https://github.com/ROCm/ROCm/issues/4082).
### Stale file due to OpenCL ICD loader deprecation
When upgrading from ROCm 6.2.x to ROCm 6.3.0, the issue of removal of the `rocm-icd-loader` package
leaving a stale file in the old `rocm-6.2.x` directory has been resolved. The stale files left during
the upgrade from ROCm 6.2.x to ROCm 6.3.0 will be removed when upgrading to ROCm 6.3.1. For more
information, see [GitHub issue #4084](https://github.com/ROCm/ROCm/issues/4084).

View File

@@ -0,0 +1,9 @@
## Operating system and hardware support changes
ROCm 6.3.1 adds support for Debian 12 (kernel: 6.1). Debian is supported only on AMD Instinct accelerators. See the installation instructions at [Debian native installation](https://rocm.docs.amd.com/projects/install-on-linux/en/docs-6.3.1/install/native-install/debian.html).
ROCm 6.3.1 enables support for AMD Instinct MI325X accelerator. For more information, see [AMD Instinct™ MI325X Accelerators](https://www.amd.com/en/products/accelerators/instinct/mi300/mi325x.html).
See the [Compatibility
matrix](https://rocm.docs.amd.com/en/docs-6.3.1/compatibility/compatibility-matrix.html)
for more information about operating system and hardware compatibility.

View File

@@ -0,0 +1,13 @@
## ROCm upcoming changes
The following changes to the ROCm software stack are anticipated for future releases.
### AMDGPU wavefront size compiler macro deprecation
The `__AMDGCN_WAVEFRONT_SIZE__` macro will be deprecated in an upcoming
release. It is recommended to remove any use of this macro. For more information, see [AMDGPU
support](https://rocm.docs.amd.com/projects/llvm-project/en/docs-6.3.1/LLVM/clang/html/AMDGPUSupport.html).
### HIPCC Perl scripts deprecation
The HIPCC Perl scripts (`hipcc.pl` and `hipconfig.pl`) will be removed in an upcoming release.

View File

@@ -66,46 +66,44 @@ endef
# It is a space seperated list with zero or more elements.
$(call adddep,amd_smi_lib,${ASAN_DEP})
$(call adddep,aqlprofile,${ASAN_DEP} hsa)
$(call adddep,aqlprofile,${ASAN_DEP} rocr)
$(call adddep,comgr,lightning devicelibs)
$(call adddep,dbgapi,hsa comgr)
$(call adddep,dbgapi,rocr comgr)
$(call adddep,devicelibs,lightning)
$(call adddep,hip_on_rocclr,${ASAN_DEP} hsa comgr hipcc rocprofiler-register)
$(call adddep,hip_on_rocclr,${ASAN_DEP} rocr comgr hipcc rocprofiler-register)
$(call adddep,hipcc,)
$(call adddep,hipify_clang,hip_on_rocclr lightning)
$(call adddep,hsa,${ASAN_DEP} thunk lightning devicelibs rocprofiler-register)
$(call adddep,lightning,)
$(call adddep,omniperf,${ASAN_DEP})
$(call adddep,omnitrace,hipcc hsa hip_on_rocclr rocm_smi_lib rocprofiler roctracer)
$(call adddep,opencl_icd_loader,)
$(call adddep,opencl_on_rocclr,${ASAN_DEP} hsa comgr opencl_icd_loader)
$(call adddep,openmp_extras,thunk lightning devicelibs hsa)
$(call adddep,rdc,${ASAN_DEP} rocm_smi_lib hsa rocprofiler)
$(call adddep,rocclr,${ASAN_DEP} hsa comgr hipcc rocprofiler-register)
$(call adddep,rocm_bandwidth_test,${ASAN_DEP} hsa)
$(call adddep,opencl_on_rocclr,${ASAN_DEP} rocr comgr)
$(call adddep,openmp_extras,lightning devicelibs rocr)
$(call adddep,rocm_bandwidth_test,${ASAN_DEP} rocr)
$(call adddep,rocm_smi_lib,${ASAN_DEP})
$(call adddep,rocm-cmake,${ASAN_DEP})
$(call adddep,rocm-core,${ASAN_DEP})
$(call adddep,rocm-gdb,dbgapi)
$(call adddep,rocminfo,${ASAN_DEP} hsa)
$(call adddep,rocminfo,${ASAN_DEP} rocr)
$(call adddep,rocprofiler-register,${ASAN_DEP})
$(call adddep,rocprofiler-sdk,${ASAN_DEP} hsa aqlprofile opencl_on_rocclr hip_on_rocclr comgr)
$(call adddep,rocprofiler,${ASAN_DEP} hsa roctracer aqlprofile opencl_on_rocclr hip_on_rocclr comgr)
$(call adddep,rocr_debug_agent,${ASAN_DEP} hip_on_rocclr hsa dbgapi)
$(call adddep,roctracer,${ASAN_DEP} hsa hip_on_rocclr)
$(call adddep,thunk,${ASAN_DEP})
$(call adddep,rocprofiler-sdk,${ASAN_DEP} rocr aqlprofile opencl_on_rocclr hip_on_rocclr comgr)
$(call adddep,rocprofiler-systems,${ASAN_DEP} hipcc rocr hip_on_rocclr rocm_smi_lib rocprofiler roctracer rocprofiler-sdk)
$(call adddep,rocprofiler,${ASAN_DEP} rocr roctracer aqlprofile opencl_on_rocclr hip_on_rocclr comgr)
$(call adddep,rocprofiler-compute,${ASAN_DEP})
$(call adddep,rocr,${ASAN_DEP} lightning rocm_smi_lib devicelibs rocprofiler-register)
$(call adddep,rocr_debug_agent,${ASAN_DEP} hip_on_rocclr rocr dbgapi)
$(call adddep,roctracer,${ASAN_DEP} rocr hip_on_rocclr)
# rocm-dev points to all possible last finish components of Stage1 build.
rocm-dev-components :=rdc hipify_clang openmp_extras \
omniperf omnitrace rocm-core amd_smi_lib hipcc \
rocm_bandwidth_test rocr_debug_agent rocm-gdb
$(call adddep,rocm-dev,$(filter-out ${NOBUILD},${rocm-dev-components}))
rocm-dev-components :=amd_smi_lib aqlprofile comgr dbgapi devicelibs hip_on_rocclr hipcc hipify_clang \
lightning rocprofiler-compute opencl_on_rocclr openmp_extras rocm_bandwidth_test rocm_smi_lib \
rocm-cmake rocm-core rocm-gdb rocminfo rocprofiler-register rocprofiler-sdk rocprofiler-systems \
rocprofiler rocr rocr_debug_agent roctracer
$(call adddep,rocm-dev,$(filter-out ${NOBUILD} kernel_ubuntu,${rocm-dev-components}))
$(call adddep,amdmigraphx,hip_on_rocclr half rocblas miopen-hip lightning hipcc)
$(call adddep,amdmigraphx,hip_on_rocclr half rocblas miopen-hip lightning hipcc hiptensor)
$(call adddep,composable_kernel,lightning hipcc hip_on_rocclr rocm-cmake)
$(call adddep,half,rocm-cmake)
$(call adddep,hipblas-common,lightning)
$(call adddep,hipblas,hip_on_rocclr rocblas rocsolver lightning hipcc)
$(call adddep,hipblaslt,hip_on_rocclr openmp_extras hipblas lightning hipcc)
$(call adddep,hipblaslt,hip_on_rocclr openmp_extras lightning hipcc hipblas-common rocm-dev)
$(call adddep,hipcub,hip_on_rocclr rocprim lightning hipcc)
$(call adddep,hipfft,hip_on_rocclr openmp_extras rocfft rocrand hiprand lightning hipcc)
$(call adddep,hipfort,rocblas hipblas rocsparse hipsparse rocfft hipfft rocrand hiprand rocsolver hipsolver lightning hipcc)
@@ -115,22 +113,25 @@ $(call adddep,hipsparse,hip_on_rocclr rocsparse lightning hipcc)
$(call adddep,hipsparselt,hip_on_rocclr hipsparse lightning hipcc openmp_extras)
$(call adddep,hiptensor,hip_on_rocclr composable_kernel lightning hipcc)
$(call adddep,miopen-deps,lightning hipcc)
$(call adddep,miopen-hip,composable_kernel half hip_on_rocclr miopen-deps rocblas roctracer lightning hipcc)
$(call adddep,miopen-hip,composable_kernel half hip_on_rocclr miopen-deps hipblas hipblaslt rocrand roctracer lightning hipcc)
$(call adddep,mivisionx,amdmigraphx miopen-hip rpp lightning hipcc)
$(call adddep,rccl,hip_on_rocclr hsa lightning hipcc rocm_smi_lib hipify_clang)
$(call adddep,rccl,rocm-core hip_on_rocclr rocr lightning hipcc rocm_smi_lib hipify_clang)
$(call adddep,rdc,rocm_smi_lib rocprofiler rocmvalidationsuite)
$(call adddep,rocalution,rocblas rocsparse rocrand lightning hipcc)
$(call adddep,rocblas,hip_on_rocclr openmp_extras lightning hipcc)
$(call adddep,rocblas,hip_on_rocclr openmp_extras lightning hipcc hipblaslt)
$(call adddep,rocal,mivisionx)
$(call adddep,rocdecode,hip_on_rocclr lightning hipcc)
$(call adddep,rocdecode,hip_on_rocclr lightning hipcc amdmigraphx)
$(call adddep,rocfft,hip_on_rocclr rocrand hiprand lightning hipcc openmp_extras)
$(call adddep,rocmvalidationsuite,hip_on_rocclr hsa rocblas rocm-core lightning hipcc rocm_smi_lib)
$(call adddep,rocjpeg,hip_on_rocclr lightning hipcc rocm-dev)
$(call adddep,rocmvalidationsuite,hip_on_rocclr rocr hipblas hiprand hipblaslt rocm-core lightning hipcc rocm_smi_lib)
$(call adddep,rocprim,hip_on_rocclr lightning hipcc)
$(call adddep,rocrand,hip_on_rocclr lightning hipcc)
$(call adddep,rocsolver,hip_on_rocclr rocblas rocsparse lightning hipcc)
$(call adddep,rocsolver,hip_on_rocclr rocblas rocsparse rocprim lightning hipcc)
$(call adddep,rocsparse,hip_on_rocclr rocprim lightning hipcc)
$(call adddep,rocthrust,hip_on_rocclr rocprim lightning hipcc)
$(call adddep,rocwmma,hip_on_rocclr rocblas lightning hipcc rocm-cmake rocm_smi_lib)
$(call adddep,rpp,half lightning hipcc openmp_extras)
$(call adddep,transferbench,hip_on_rocclr lightning hipcc)
# -------------------------------------------------------------------------
@@ -189,7 +190,7 @@ else # } {
# Pass in jobserver info using the RMAKE variable
${RMAKE}@( if set -x && source $${INFRA_REPO}/envsetup.sh && \
rm -f $$@.errors $$@ $$@.repackaged && \
$${INFRA_REPO}/build_$1.sh -c && source $${INFRA_REPO}/ccache-env-mathlib.sh && \
$${INFRA_REPO}/build_$1.sh -c && \
time bash -x $${INFRA_REPO}/build_$1.sh $${RELEASE_FLAG} $${SANITIZER_FLAG} && $${INFRA_REPO}/post_inst_pkg.sh "$1" ; \
then mv $$@.inprogress $$@ ; \
else mv $$@.inprogress $$@.errors ; echo Error in $1 >&2 ; exit 1 ;\
@@ -216,11 +217,14 @@ $(call peval,$(foreach dep,$(strip ${components}),$(call toplevel,${dep})))
all: $(addprefix T_,$(filter-out ${NOBUILD},${components}))
@echo All ROCm components built
# Do not document this target
upload: $(addprefix U_,${components})
upload: $(addprefix U_,$(filter-out ${NOBUILD},${components}))
@echo All ROCm components built and uploaded
upload-rocm-dev: $(addprefix U_,$(filter-out ${NOBUILD},${components}))
@echo All rocm-dev components built and uploaded
##help rocm-dev: Build a subset of ROCm
rocm-dev: T_rocm-dev
rocm-dev: $(addprefix T_,$(filter-out ${NOBUILD},${components}))
@echo rocm-dev built
${OUT_DIR}/logs:

View File

@@ -22,15 +22,15 @@ printUsage() {
return 0
}
PROJ_NAME="amdsmi"
PACKAGE_ROOT="$(getPackageRoot)"
TARGET="build"
PACKAGE_LIB=$(getLibPath)
PACKAGE_INCLUDE="$(getIncludePath)"
AMDSMI_BUILD_DIR=$(getBuildPath amdsmi)
AMDSMI_PACKAGE_DEB_DIR="$(getPackageRoot)/deb/amdsmi"
AMDSMI_PACKAGE_RPM_DIR="$(getPackageRoot)/rpm/amdsmi"
AMDSMI_BUILD_DIR=$(getBuildPath $PROJ_NAME)
AMDSMI_PACKAGE_DEB_DIR="$PACKAGE_ROOT/deb/$PROJ_NAME"
AMDSMI_PACKAGE_RPM_DIR="$PACKAGE_ROOT/rpm/$PROJ_NAME"
AMDSMI_BUILD_TYPE="debug"
BUILD_TYPE="Debug"
@@ -57,10 +57,9 @@ do
(-a | --address_sanitizer)
set_asan_env_vars
set_address_sanitizer_on
# TODO - support standard option of passing cmake environment vars - CFLAGS,CXXFLAGS etc., to enable address sanitizer
ADDRESS_SANITIZER=true ; shift ;;
(-s | --static)
SHARED_LIBS="OFF" ; shift ;;
ack_and_skip_static ;;
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 2 ;;
(-p | --package)

View File

@@ -10,7 +10,9 @@ build_amdmigraphx() {
cd $COMPONENT_SRC
pip3 install https://github.com/RadeonOpenCompute/rbuild/archive/master.tar.gz
if ! command -v rbuild &> /dev/null; then
pip3 install https://github.com/RadeonOpenCompute/rbuild/archive/master.tar.gz
fi
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
@@ -20,7 +22,7 @@ build_amdmigraphx() {
if [ -n "$GPU_ARCHS" ]; then
GPU_TARGETS="$GPU_ARCHS"
else
GPU_TARGETS="gfx908;gfx90a;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101"
GPU_TARGETS="gfx900;gfx906;gfx908;gfx90a;gfx1030;gfx1100;gfx1101;gfx1102;gfx942;gfx1200;gfx1201"
fi
init_rocm_common_cmake_params
@@ -29,7 +31,7 @@ build_amdmigraphx() {
--cxx="${ROCM_PATH}/llvm/bin/clang++" \
--cc="${ROCM_PATH}/llvm/bin/clang" \
"${rocm_math_common_cmake_params[@]}" \
-DCMAKE_MODULE_LINKER_FLAGS="-Wl,--enable-new-dtags -Wl,--rpath,$ROCM_LIB_RPATH" \
-DCMAKE_MODULE_LINKER_FLAGS="-Wl,--enable-new-dtags,--build-id=sha1,--rpath,$ROCM_LIB_RPATH" \
-DGPU_TARGETS="${GPU_TARGETS}" \
-DCMAKE_INSTALL_RPATH=""

View File

@@ -11,7 +11,9 @@ printUsage() {
echo " -p, --package <type> Specify packaging format"
echo " -r, --release Make a release build instead of a debug build"
echo " -o, --outdir <pkg_type> Print path of output directory containing packages of
type referred to by pkg_type"
type referred to by pkg_type"
echo " -s, --static Component/Build does not support static builds just accepting this param & ignore.
No effect of the param on this build"
echo " -h, --help Prints this help"
echo
echo "Possible values for <type>:"

View File

@@ -1,136 +0,0 @@
#!/bin/bash
source "$(dirname "${BASH_SOURCE}")/compute_utils.sh"
printUsage() {
echo
echo "Usage: $(basename "${BASH_SOURCE}") [-c|-r|-h] [makeopts]"
echo
echo "Options:"
echo " -c, --clean Removes all clang-ocl build artifacts"
echo " -r, --release Build non-debug version clang-ocl (default is debug)"
echo " -a, --address_sanitizer Enable address sanitizer"
echo " -o, --outdir <pkg_type> Print path of output directory containing packages of
type referred to by pkg_type"
echo " -h, --help Prints this help"
echo " -s, --static Supports static CI by accepting this param & not bailing out. No effect of the param though"
echo
return 0
}
TARGET="build"
CLANG_OCL_DEST="$(getBinPath)"
CLANG_OCL_SRC_ROOT="$CLANG_OCL_ROOT"
CLANG_OCL_BUILD_DIR="$(getBuildPath clang-ocl)"
MAKEARG="$DASH_JAY"
PACKAGE_ROOT="$(getPackageRoot)"
PACKAGE_UTILS="$(getUtilsPath)"
CLANG_OCL_PACKAGE_DEB="$PACKAGE_ROOT/deb/clang-ocl"
CLANG_OCL_PACKAGE_RPM="$PACKAGE_ROOT/rpm/clang-ocl"
BUILD_TYPE="Debug"
SHARED_LIBS="ON"
CLEAN_OR_OUT=0;
MAKETARGET="deb"
PKGTYPE="deb"
VALID_STR=`getopt -o hcraso:g: --long help,clean,release,clean,static,address_sanitizer,outdir:,gpu_list: -- "$@"`
eval set -- "$VALID_STR"
while true ;
do
case "$1" in
(-h | --help)
printUsage ; exit 0;;
(-c | --clean)
TARGET="clean" ; ((CLEAN_OR_OUT|=1)) ; shift ;;
(-r | --release)
MAKEARG="$MAKEARG BUILD_TYPE=rel" ; BUILD_TYPE="Release" ; shift ;;
(-a | --address_sanitizer)
set_asan_env_vars
set_address_sanitizer_on ; shift ;;
(-s | --static)
SHARED_LIBS="OFF" ; shift ;;
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 2 ;;
(-g | --gpu_list )
GPU_LIST=$2; shift 2 ;;
--) shift; break;;
(*)
echo " This should never come but just incase : UNEXPECTED ERROR Parm : [$1] ">&2 ; exit 20;;
esac
done
RET_CONFLICT=1
check_conflicting_options $CLEAN_OR_OUT $PKGTYPE $MAKETARGET
if [ $RET_CONFLICT -ge 30 ]; then
print_vars $API_NAME $TARGET $BUILD_TYPE $SHARED_LIBS $CLEAN_OR_OUT $PKGTYPE $MAKETARGET
exit $RET_CONFLICT
fi
clean_clang-ocl() {
echo "Removing clang-ocl"
rm -rf $CLANG_OCL_DEST/clang-ocl
rm -rf $CLANG_OCL_BUILD_DIR
rm -rf $CLANG_OCL_PACKAGE_DEB
rm -rf $CLANG_OCL_PACKAGE_RPM
}
build_clang-ocl() {
if [ ! -d "$CLANG_OCL_BUILD_DIR" ]; then
mkdir -p $CLANG_OCL_BUILD_DIR
pushd $CLANG_OCL_BUILD_DIR
if [ -e $PACKAGE_ROOT/lib/bitcode/opencl.amdgcn.bc ]; then
BC_DIR="$ROCM_INSTALL_PATH/lib"
else
BC_DIR="$ROCM_INSTALL_PATH/amdgcn/bitcode"
fi
cmake \
$(rocm_cmake_params) \
-DDISABLE_CHECKS="ON" \
-DCLANG_BIN="$ROCM_INSTALL_PATH/llvm/bin" \
-DBITCODE_DIR="$BC_DIR" \
$(rocm_common_cmake_params) \
-DCPACK_SET_DESTDIR="OFF" \
$CLANG_OCL_SRC_ROOT
echo "Making clang-ocl:"
cmake --build . -- $MAKEARG
cmake --build . -- $MAKEARG install
cmake --build . -- $MAKEARG package
popd
fi
copy_if DEB "${CPACKGEN:-"DEB;RPM"}" "$CLANG_OCL_PACKAGE_DEB" $CLANG_OCL_BUILD_DIR/rocm-clang-ocl*.deb
copy_if RPM "${CPACKGEN:-"DEB;RPM"}" "$CLANG_OCL_PACKAGE_RPM" $CLANG_OCL_BUILD_DIR/rocm-clang-ocl*.rpm
}
print_output_directory() {
case ${PKGTYPE} in
("deb")
echo ${CLANG_OCL_PACKAGE_DEB};;
("rpm")
echo ${CLANG_OCL_PACKAGE_RPM};;
(*)
echo "Invalid package type \"${PKGTYPE}\" provided for -o" >&2; exit 1;;
esac
exit
}
case $TARGET in
(clean) clean_clang-ocl ;;
(build) build_clang-ocl ;;
(outdir) print_output_directory ;;
(*) die "Invalid target $TARGET" ;;
esac
echo "Operation complete"
exit 0

View File

@@ -6,73 +6,53 @@ source "$(dirname "${BASH_SOURCE[0]}")/compute_helper.sh"
set_component_src composable_kernel
GPU_ARCH_LIST="gfx908;gfx90a;gfx942;gfx1030;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201"
build_miopen_ck() {
echo "Start Building Composable Kernel"
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
set_address_sanitizer_on
GPU_ARCH_LIST="gfx908:xnack+;gfx90a:xnack+;gfx942:xnack+"
else
unset_asan_env_vars
set_address_sanitizer_off
fi
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
GPU_ARCH_LIST="gfx942"
ack_and_skip_static
fi
PYTHON_VERSION_WORKAROUND=''
echo "DISTRO_ID: ${DISTRO_ID}"
if [ "$DISTRO_ID" = "rhel-8.8" ] || [ "$DISTRO_ID" = "sles-15.5" ] ; then
EXTRA_PYTHON_PATH=/opt/Python-3.8.13
PYTHON_VERSION_WORKAROUND="-DCK_USE_ALTERNATIVE_PYTHON=${EXTRA_PYTHON_PATH}/bin/python3.8"
# For the python interpreter we need to export LD_LIBRARY_PATH.
export LD_LIBRARY_PATH=${EXTRA_PYTHON_PATH}/lib:$LD_LIBRARY_PATH
fi
cd $COMPONENT_SRC
mkdir "$BUILD_DIR" && cd "$BUILD_DIR"
init_rocm_common_cmake_params
if [ -n "$GPU_ARCHS" ]; then
GPU_TARGETS="-DAMDGPU_TARGETS=${GPU_ARCHS}"
fi
if [ "${ASAN_CMAKE_PARAMS}" == "true" ] ; then
cmake -DBUILD_DEV=OFF \
-DCMAKE_BUILD_TYPE=${BUILD_TYPE:-'RelWithDebInfo'} \
-DCMAKE_CXX_COMPILER=${ROCM_PATH}/llvm/bin/clang++ \
-DCMAKE_CXX_FLAGS=" -O3 " \
-DCMAKE_PREFIX_PATH="${ROCM_PATH%-*}/lib/cmake;${ROCM_PATH%-*}/$ASAN_LIBDIR;${ROCM_PATH%-*}/llvm;${ROCM_PATH%-*}" \
-DCMAKE_SHARED_LINKER_FLAGS_INIT="-Wl,--enable-new-dtags,--rpath,$ROCM_ASAN_LIB_RPATH" \
-DCMAKE_EXE_LINKER_FLAGS_INIT="-Wl,--enable-new-dtags,--rpath,$ROCM_ASAN_EXE_RPATH" \
-DCMAKE_VERBOSE_MAKEFILE=1 \
-DCMAKE_INSTALL_RPATH_USE_LINK_PATH=FALSE \
-DCMAKE_INSTALL_PREFIX=${ROCM_PATH} \
-DCMAKE_PACKAGING_INSTALL_PREFIX=${ROCM_PATH} \
-DBUILD_FILE_REORG_BACKWARD_COMPATIBILITY=OFF \
-DROCM_SYMLINK_LIBS=OFF \
-DCPACK_PACKAGING_INSTALL_PREFIX=${ROCM_PATH} \
-DROCM_DISABLE_LDCONFIG=ON \
-DROCM_PATH=${ROCM_PATH} \
-DCPACK_GENERATOR="${PKGTYPE^^}" \
${LAUNCHER_FLAGS} \
-DINSTANCES_ONLY=ON \
-DENABLE_ASAN_PACKAGING=true \
"${GPU_TARGETS}" \
"$COMPONENT_SRC"
else
cmake -DBUILD_DEV=OFF \
-DCMAKE_BUILD_TYPE=Release \
-DCMAKE_CXX_COMPILER=${ROCM_PATH}/llvm/bin/clang++ \
-DCMAKE_CXX_FLAGS=" -O3 " \
-DCMAKE_PREFIX_PATH=${ROCM_PATH%-*} \
-DCMAKE_SHARED_LINKER_FLAGS_INIT='-Wl,--enable-new-dtags,--rpath,$ORIGIN' \
-DCMAKE_EXE_LINKER_FLAGS_INIT='-Wl,--enable-new-dtags,--rpath,$ORIGIN/../lib' \
-DCMAKE_VERBOSE_MAKEFILE=1 \
-DCMAKE_INSTALL_RPATH_USE_LINK_PATH=FALSE \
-DCMAKE_INSTALL_PREFIX=${ROCM_PATH} \
-DCMAKE_PACKAGING_INSTALL_PREFIX=${ROCM_PATH} \
-DBUILD_FILE_REORG_BACKWARD_COMPATIBILITY=OFF \
-DROCM_SYMLINK_LIBS=OFF \
-DCPACK_PACKAGING_INSTALL_PREFIX=${ROCM_PATH} \
-DROCM_DISABLE_LDCONFIG=ON \
-DROCM_PATH=${ROCM_PATH} \
-DCPACK_GENERATOR="${PKGTYPE^^}" \
-DCMAKE_CXX_COMPILER="${ROCM_PATH}/llvm/bin/clang++" \
-DCMAKE_C_COMPILER="${ROCM_PATH}/llvm/bin/clang" \
${LAUNCHER_FLAGS} \
-DINSTANCES_ONLY=ON \
"${GPU_TARGETS}" \
"$COMPONENT_SRC"
fi
cmake \
-DBUILD_DEV=OFF \
"${rocm_math_common_cmake_params[@]}" \
${PYTHON_VERSION_WORKAROUND} \
-DCPACK_GENERATOR="${PKGTYPE^^}" \
-DCMAKE_CXX_COMPILER="${ROCM_PATH}/llvm/bin/clang++" \
-DCMAKE_C_COMPILER="${ROCM_PATH}/llvm/bin/clang" \
${LAUNCHER_FLAGS} \
-DGPU_ARCHS="${GPU_ARCH_LIST}" \
-DCMAKE_BUILD_TYPE=Release \
-DCMAKE_CXX_FLAGS=" -O3 " \
"$COMPONENT_SRC"
cmake --build . -- -j${PROC} package
cmake --build "$BUILD_DIR" -- install
mkdir -p $PACKAGE_DIR && cp ./*.${PKGTYPE} $PACKAGE_DIR
rm -rf *
}
unset_asan_env_vars() {
@@ -88,85 +68,6 @@ set_address_sanitizer_off() {
export LDFLAGS=""
}
build_miopen_ckProf() {
ENABLE_ADDRESS_SANITIZER=false
echo "Start Building Composable Kernel Profiler"
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
set_address_sanitizer_on
else
unset_asan_env_vars
set_address_sanitizer_off
fi
cd $COMPONENT_SRC
cd "$BUILD_DIR"
rm -rf *
architectures='gfx10 gfx11 gfx90 gfx94'
if [ -n "$GPU_ARCHS" ]; then
architectures=$(echo ${GPU_ARCHS} | awk -F';' '{for(i=1;i<=NF;i++) a[substr($i,1,5)]} END{for(i in a) printf i" "}')
fi
for arch in ${architectures}
do
if [ "${ASAN_CMAKE_PARAMS}" == "true" ] ; then
cmake -DBUILD_DEV=OFF \
-DCMAKE_PREFIX_PATH="${ROCM_PATH%-*}/lib/cmake;${ROCM_PATH%-*}/$ASAN_LIBDIR;${ROCM_PATH%-*}/llvm;${ROCM_PATH%-*}" \
-DCMAKE_BUILD_TYPE=${BUILD_TYPE:-'RelWithDebInfo'} \
-DCMAKE_SHARED_LINKER_FLAGS_INIT="-Wl,--enable-new-dtags,--rpath,$ROCM_ASAN_LIB_RPATH" \
-DCMAKE_EXE_LINKER_FLAGS_INIT="-Wl,--enable-new-dtags,--rpath,$ROCM_ASAN_EXE_RPATH" \
-DCMAKE_VERBOSE_MAKEFILE=1 \
-DCMAKE_INSTALL_RPATH_USE_LINK_PATH=FALSE \
-DCMAKE_INSTALL_PREFIX="${ROCM_PATH}" \
-DCMAKE_PACKAGING_INSTALL_PREFIX="${ROCM_PATH}" \
-DBUILD_FILE_REORG_BACKWARD_COMPATIBILITY=OFF \
-DROCM_SYMLINK_LIBS=OFF \
-DCPACK_PACKAGING_INSTALL_PREFIX="${ROCM_PATH}" \
-DROCM_DISABLE_LDCONFIG=ON \
-DROCM_PATH="${ROCM_PATH}" \
-DCPACK_GENERATOR="${PKGTYPE^^}" \
-DCMAKE_CXX_COMPILER="${ROCM_PATH}/llvm/bin/clang++" \
-DCMAKE_C_COMPILER="${ROCM_PATH}/llvm/bin/clang" \
${LAUNCHER_FLAGS} \
-DPROFILER_ONLY=ON \
-DENABLE_ASAN_PACKAGING=true \
-DGPU_ARCH="${arch}" \
"$COMPONENT_SRC"
else
cmake -DBUILD_DEV=OFF \
-DCMAKE_PREFIX_PATH="${ROCM_PATH%-*}" \
-DCMAKE_BUILD_TYPE=Release \
-DCMAKE_SHARED_LINKER_FLAGS_INIT='-Wl,--enable-new-dtags,--rpath,$ORIGIN' \
-DCMAKE_EXE_LINKER_FLAGS_INIT='-Wl,--enable-new-dtags,--rpath,$ORIGIN/../lib' \
-DCMAKE_VERBOSE_MAKEFILE=1 \
-DCMAKE_INSTALL_RPATH_USE_LINK_PATH=FALSE \
-DCMAKE_INSTALL_PREFIX="${ROCM_PATH}" \
-DCMAKE_PACKAGING_INSTALL_PREFIX="${ROCM_PATH}" \
-DBUILD_FILE_REORG_BACKWARD_COMPATIBILITY=OFF \
-DROCM_SYMLINK_LIBS=OFF \
-DCPACK_PACKAGING_INSTALL_PREFIX="${ROCM_PATH}" \
-DROCM_DISABLE_LDCONFIG=ON \
-DROCM_PATH="${ROCM_PATH}" \
-DCPACK_GENERATOR="${PKGTYPE^^}" \
-DCMAKE_CXX_COMPILER="${ROCM_PATH}/llvm/bin/clang++" \
-DCMAKE_C_COMPILER="${ROCM_PATH}/llvm/bin/clang" \
${LAUNCHER_FLAGS} \
-DPROFILER_ONLY=ON \
-DGPU_ARCH="${arch}" \
"$COMPONENT_SRC"
fi
cmake --build . -- -j${PROC} package
cp ./*ckprofiler*.${PKGTYPE} $PACKAGE_DIR
rm -rf *
done
rm -rf _CPack_Packages/ && find -name '*.o' -delete
echo "Finished building Composable Kernel"
show_build_cache_stats
}
clean_miopen_ck() {
echo "Cleaning MIOpen-CK build directory: ${BUILD_DIR} ${PACKAGE_DIR}"
rm -rf "$BUILD_DIR" "$PACKAGE_DIR"
@@ -176,7 +77,7 @@ clean_miopen_ck() {
stage2_command_args "$@"
case $TARGET in
build) build_miopen_ck; build_miopen_ckProf;;
build) build_miopen_ck ;;
outdir) print_output_directory ;;
clean) clean_miopen_ck ;;
*) die "Invalid target $TARGET" ;;

View File

@@ -15,7 +15,7 @@ printUsage() {
type referred to by pkg_type"
echo " -h, --help Prints this help"
echo " -M, --skip_man_pages Do not build the 'docs' target"
echo " -s, --static Supports static CI by accepting this param & not bailing out. No effect of the param though"
echo " -s, --static Component/Build does not support static builds just accepting this param & ignore. No effect of the param on this build"
echo
echo "Possible values for <type>:"
echo " deb -> Debian format (default)"
@@ -65,7 +65,7 @@ do
set_asan_env_vars
set_address_sanitizer_on ;;
(-s | --static)
SHARED_LIBS="OFF" ;;
ack_and_skip_static ;;
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; ((CLEAN_OR_OUT|=2)) ; shift 1 ;;
(-M | --skip_man_pages) DODOCSBUILD=false;;

View File

@@ -96,6 +96,7 @@ build_devicelibs() {
if [ ! -e Makefile ]; then
cmake $(rocm_cmake_params) \
$(rocm_common_cmake_params) \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
-DROCM_DEVICE_LIBS_BITCODE_INSTALL_LOC_NEW="$bitcodeInstallLoc/amdgcn" \
-DROCM_DEVICE_LIBS_BITCODE_INSTALL_LOC_OLD="amdgcn" \
"$DEVICELIBS_ROOT"

View File

@@ -24,14 +24,15 @@ printUsage() {
source "$(dirname "${BASH_SOURCE}")/compute_utils.sh"
MAKEOPTS="$DASH_JAY"
PROJ_NAME="hip-on-rocclr"
BUILD_PATH="$(getBuildPath hip-on-rocclr)"
BUILD_PATH="$(getBuildPath $PROJ_NAME)"
TARGET="build"
PACKAGE_ROOT="$(getPackageRoot)"
PACKAGE_SRC="$(getSrcPath)"
PACKAGE_DEB="$PACKAGE_ROOT/deb/hip-on-rocclr"
PACKAGE_RPM="$PACKAGE_ROOT/rpm/hip-on-rocclr"
PACKAGE_DEB="$PACKAGE_ROOT/deb/$PROJ_NAME"
PACKAGE_RPM="$PACKAGE_ROOT/rpm/$PROJ_NAME"
PREFIX_PATH="$PACKAGE_ROOT"
CORE_BUILD_DIR="$(getBuildPath hsa-core)"
ROCclr_BUILD_DIR="$(getBuildPath rocclr)"
@@ -52,7 +53,7 @@ MAKETARGET="deb"
PKGTYPE="deb"
OFFLOAD_ARCH=()
DEFAULT_OFFLOAD_ARCH=(gfx900 gfx906 gfx908 gfx90a gfx940 gfx941 gfx942 gfx1030 gfx1031 gfx1033 gfx1034 gfx1035 gfx1100 gfx1101 gfx1102 gfx1103)
DEFAULT_OFFLOAD_ARCH=(gfx900 gfx906 gfx908 gfx90a gfx940 gfx941 gfx942 gfx1030 gfx1031 gfx1033 gfx1034 gfx1035 gfx1100 gfx1101 gfx1102 gfx1200 gfx1201)
VALID_STR=`getopt -o hcrast:o: --long help,clean,release,address_sanitizer,static,offload-arch=:,outdir: -- "$@"`
eval set -- "$VALID_STR"
@@ -168,9 +169,11 @@ build_catch_tests() {
export ROCM_PATH="$ROCM_INSTALL_PATH"
cmake \
-DCMAKE_BUILD_TYPE="${BUILD_TYPE}" \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
-DHIP_PLATFORM=amd \
-DROCM_PATH="$ROCM_INSTALL_PATH" \
-DOFFLOAD_ARCH_STR="$OFFLOAD_ARCH_STR" \
$(rocm_cmake_params) \
$(rocm_common_cmake_params) \
-DCPACK_RPM_DEBUGINFO_PACKAGE=FALSE \
-DCPACK_DEBIAN_DEBUGINFO_PACKAGE=FALSE \
@@ -206,6 +209,8 @@ package_samples() {
export ROCM_PATH="$ROCM_INSTALL_PATH"
cmake \
-DROCM_PATH="$ROCM_INSTALL_PATH" \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
$(rocm_cmake_params) \
$(rocm_common_cmake_params) \
-DCMAKE_MODULE_PATH="$CMAKE_PATH/hip" \
-DCPACK_INSTALL_PREFIX="$ROCM_INSTALL_PATH" \

View File

@@ -0,0 +1,41 @@
#!/bin/bash
set -ex
source "$(dirname "${BASH_SOURCE[0]}")/compute_helper.sh"
set_component_src hipBLAS-common
build_hipblas-common() {
echo "Start build"
cd $COMPONENT_SRC
mkdir -p "$BUILD_DIR" && cd "$BUILD_DIR"
init_rocm_common_cmake_params
cmake \
"${rocm_math_common_cmake_params[@]}" \
"$COMPONENT_SRC"
cmake --build "$BUILD_DIR" -- install
cmake --build "$BUILD_DIR" -- package
rm -rf _CPack_Packages/ && find -name '*.o' -delete
mkdir -p $PACKAGE_DIR && cp ${BUILD_DIR}/*.${PKGTYPE} $PACKAGE_DIR
show_build_cache_stats
}
clean_hipblas-common() {
echo "Cleaning hipBLAS-common build directory: ${BUILD_DIR} ${PACKAGE_DIR}"
rm -rf "$BUILD_DIR" "$PACKAGE_DIR"
echo "Done!"
}
stage2_command_args "$@"
case $TARGET in
build) build_hipblas-common ;;
outdir) print_output_directory ;;
clean) clean_hipblas-common ;;
*) die "Invalid target $TARGET" ;;
esac

View File

@@ -10,6 +10,12 @@ build_hipblas() {
echo "Start build"
CXX="g++"
CXX_FLAG=
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
CXX="amdclang++"
CXX_FLAG="-DCMAKE_CXX_COMPILER=${ROCM_PATH}/llvm/bin/clang++"
fi
CLIENTS_SAMPLES="ON"
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
@@ -17,6 +23,8 @@ build_hipblas() {
CLIENTS_SAMPLES="OFF"
fi
SHARED_LIBS="ON"
echo "C compiler: $CC"
echo "CXX compiler: $CXX"
echo "FC compiler: $FC"
@@ -33,11 +41,12 @@ build_hipblas() {
${LAUNCHER_FLAGS} \
"${rocm_math_common_cmake_params[@]}" \
-DUSE_CUDA=OFF \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
-DBUILD_CLIENTS_TESTS=ON \
-DBUILD_CLIENTS_BENCHMARKS=ON \
-DBUILD_CLIENTS_SAMPLES="${CLIENTS_SAMPLES}" \
-DCPACK_SET_DESTDIR=OFF \
-DBUILD_ADDRESS_SANITIZER="${ADDRESS_SANITIZER}" \
${CXX_FLAG} \
"$COMPONENT_SRC"
cmake --build "$BUILD_DIR" -- -j${PROC}

View File

@@ -8,6 +8,10 @@ set_component_src hipBLASLt
build_hipblaslt() {
echo "Start build"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
ack_and_skip_static
fi
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
set_address_sanitizer_on
@@ -40,7 +44,6 @@ build_hipblaslt() {
-DBUILD_CLIENTS_SAMPLES=ON \
-DBUILD_CLIENTS_TESTS=ON \
-DBUILD_CLIENTS_BENCHMARKS=ON \
-DCPACK_SET_DESTDIR=OFF \
-DBUILD_ADDRESS_SANITIZER="${ADDRESS_SANITIZER}" \
"$COMPONENT_SRC"

View File

@@ -9,10 +9,12 @@ printUsage() {
echo "Options:"
echo " -a, --address_sanitizer Enable address sanitizer"
echo " -c, --clean Clean output and delete all intermediate work"
echo " -h, --help Prints this help"
echo " -o, --outdir <pkg_type> Print path of output directory containing packages of
type referred to by pkg_type"
echo " -r, --release Makes a release build"
echo " -h, --help Prints this help"
echo
echo " -s, --static Build static lib (.a). build instead of dynamic/shared(.so) "
echo
return 0
@@ -25,25 +27,31 @@ PROJ_NAME=$API_NAME
TARGET="build"
MAKEOPTS="$DASH_JAY"
BUILD_TYPE="Debug"
SHARED_LIBS="ON"
BUILD_DIR=$(getBuildPath $API_NAME)
PACKAGE_DEB=$(getPackageRoot)/deb/$API_NAME
PACKAGE_RPM=$(getPackageRoot)/rpm/$API_NAME
PACKAGE_SRC="$(getSrcPath)"
while [ "$1" != "" ];
VALID_STR=`getopt -o hcraswo:p: --long help,clean,release,address_sanitizer,static,outdir,wheel:,package: -- "$@"`
eval set -- "$VALID_STR"
while true ;
do
case $1 in
case "$1" in
(-a | --address_sanitizer)
ack_and_ignore_asan ;;
(-c | --clean)
TARGET="clean" ;;
(-o | --outdir)
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 1 ;;
(-r | --release)
BUILD_TYPE="RelWithDebInfo" ;;
(-s | --static)
SHARED_LIBS="OFF" ;;
(-h | --help)
printUsage ; exit 0 ;;
--) shift; break;;
(*)
echo "Invalid option [$1]" >&2; printUsage; exit 1 ;;
esac
@@ -79,6 +87,7 @@ build() {
fi
cmake \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
$(rocm_cmake_params) \
$(rocm_common_cmake_params) \
-DHIPCC_BACKWARD_COMPATIBILITY=OFF \
@@ -87,7 +96,7 @@ build() {
popd
cmake --build "$BUILD_DIR" -- $MAKEOPTS
echo "Installing and Packaging hipcc"
cmake --build "$BUILD_DIR" -- $MAKEOPTS install
cmake --build "$BUILD_DIR" -- $MAKEOPTS package

View File

@@ -9,6 +9,10 @@ set_component_src hipCUB
build_hipcub() {
echo "Start build"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
ack_and_skip_static
fi
cd $COMPONENT_SRC
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
@@ -22,7 +26,7 @@ build_hipcub() {
if [ -n "$GPU_ARCHS" ]; then
GPU_TARGETS="$GPU_ARCHS"
else
GPU_TARGETS="gfx908:xnack-;gfx90a:xnack-;gfx90a:xnack+;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101"
GPU_TARGETS="gfx908:xnack-;gfx90a:xnack-;gfx90a:xnack+;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201"
fi
CXX=$(set_build_variables CXX)\

View File

@@ -21,7 +21,7 @@ build_hipfft() {
if [ -n "$GPU_ARCHS" ]; then
GPU_TARGETS="$GPU_ARCHS"
else
GPU_TARGETS="gfx908;gfx90a;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101"
GPU_TARGETS="gfx908;gfx90a;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201"
fi
cmake \

View File

@@ -8,11 +8,18 @@ set_component_src hipfort
build_hipfort() {
echo "Start build"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
ack_and_skip_static
fi
mkdir -p "$BUILD_DIR" && cd "$BUILD_DIR"
cmake --trace \
cmake \
-DCPACK_PACKAGING_INSTALL_PREFIX=${ROCM_PATH}\
-DHIPFORT_INSTALL_DIR="${ROCM_PATH}" \
-DCMAKE_PREFIX_PATH="${ROCM_PATH}/llvm;${ROCM_PATH}" \
-DCMAKE_BUILD_TYPE=Release \
-DCPACK_SET_DESTDIR="OFF" \
-DCPACK_RPM_PACKAGE_RELOCATABLE="ON" \
-DHIPFORT_COMPILER="${ROCM_PATH}/${ROCM_LLVMDIR}/bin/flang" \
-DCMAKE_Fortran_FLAGS="-Mfree" \
-DHIPFORT_COMPILER_FLAGS="-cpp" \

View File

@@ -22,12 +22,12 @@ printUsage() {
TARGET="build"
MAKEOPTS="$DASH_JAY"
HIPIFY_CLANG_BUILD_DIR="$(getBuildPath $HIPIFY_ROOT)"
HIPIFY_CLANG_DIST_DIR="$HIPIFY_CLANG_BUILD_DIR/dist"
BUILD_TYPE="Debug"
PACKAGE_ROOT="$(getPackageRoot)"
HIPIFY_CLANG_HASH=""
LIGHTNING_PATH="$ROCM_INSTALL_PATH/llvm"
ADDRESS_SANITIZER=false
INSTALL_CLANG_HEADERS="OFF"
DEB_PATH="$(getDebPath hipify)"
RPM_PATH="$(getRpmPath hipify)"
SHARED_LIBS="ON"
@@ -53,7 +53,7 @@ do
set_address_sanitizer_on
ADDRESS_SANITIZER=true ; shift ;;
(-s | --static)
SHARED_LIBS="OFF" ; shift ;;
ack_and_skip_static ;;
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 2 ;;
--) shift; break;;
@@ -74,7 +74,6 @@ fi
clean_hipify() {
echo "Cleaning hipify-clang"
rm -rf "$HIPIFY_CLANG_BUILD_DIR"
rm -rf "$HIPIFY_CLANG_DIST_DIR"
rm -rf "$DEB_PATH"
rm -rf "$RPM_PATH"
}
@@ -101,16 +100,16 @@ package_hipify() {
build_hipify() {
echo "Building hipify-clang binaries"
mkdir -p "$HIPIFY_CLANG_BUILD_DIR"
mkdir -p "$HIPIFY_CLANG_DIST_DIR"
pushd "$HIPIFY_CLANG_BUILD_DIR"
cmake \
-DCMAKE_BUILD_TYPE="$BUILD_TYPE" \
$(rocm_common_cmake_params) \
-DCMAKE_INSTALL_PREFIX="$HIPIFY_CLANG_DIST_DIR" \
-DCMAKE_INSTALL_PREFIX="$ROCM_INSTALL_PATH" \
-DCPACK_PACKAGING_INSTALL_PREFIX=$ROCM_INSTALL_PATH \
-DCMAKE_PREFIX_PATH="$LIGHTNING_PATH" \
-DADDRESS_SANITIZER="$ADDRESS_SANITIZER" \
-DHIPIFY_INSTALL_CLANG_HEADERS="$INSTALL_CLANG_HEADERS" \
$HIPIFY_ROOT
cmake --build . -- $MAKEOPTS install

View File

@@ -21,6 +21,11 @@ done
build_hiprand() {
echo "Start build"
SHARED_LIBS="ON"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
SHARED_LIBS="OFF"
fi
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
set_address_sanitizer_on
@@ -34,17 +39,20 @@ build_hiprand() {
mkdir "$BUILD_DIR" && cd "$BUILD_DIR"
init_rocm_common_cmake_params
if [ -n "$GPU_ARCHS" ]; then
GPU_TARGETS="$GPU_ARCHS"
else
GPU_TARGETS="gfx908:xnack-;gfx90a:xnack-;gfx90a:xnack+;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101"
GPU_TARGETS="gfx908:xnack-;gfx90a:xnack-;gfx90a:xnack+;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201"
fi
CXX=$(set_build_variables CXX)\
cmake \
${LAUNCHER_FLAGS} \
$(rocm_common_cmake_params) \
"${rocm_math_common_cmake_params[@]}" \
-DAMDGPU_TARGETS=${GPU_TARGETS} \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
-DBUILD_TEST=ON \
-DBUILD_BENCHMARK=ON \
-DBUILD_CRUSH_TEST=ON \
@@ -60,7 +68,6 @@ build_hiprand() {
rm -rf _CPack_Packages/ && find -name '*.o' -delete
mkdir -p $PACKAGE_DIR && cp ${BUILD_DIR}/*.${PKGTYPE} $PACKAGE_DIR
}
clean_hiprand() {

View File

@@ -9,14 +9,23 @@ set_component_src hipSOLVER
build_hipsolver() {
echo "Start build"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
CXX_FLAG="-DCMAKE_CXX_COMPILER=${ROCM_PATH}/llvm/bin/clang++"
fi
cd $COMPONENT_SRC
CXX="g++"
CXX="amdclang++"
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
set_address_sanitizer_on
fi
SHARED_LIBS="ON"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
SHARED_LIBS="OFF"
fi
echo "C compiler: $CC"
echo "CXX compiler: $CXX"
echo "FC compiler: $FC"
@@ -30,13 +39,15 @@ build_hipsolver() {
init_rocm_common_cmake_params
cmake \
-DUSE_CUDA=OFF \
-DCMAKE_CXX_COMPILER=${CXX} \
${LAUNCHER_FLAGS} \
"${rocm_math_common_cmake_params[@]}" \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
-DBUILD_CLIENTS_TESTS=ON \
-DBUILD_CLIENTS_BENCHMARKS=ON \
-DBUILD_CLIENTS_SAMPLES=ON \
-DCPACK_SET_DESTDIR=OFF \
-DBUILD_ADDRESS_SANITIZER="${ADDRESS_SANITIZER}" \
${CXX_FLAG} \
"$COMPONENT_SRC"
cmake --build "$BUILD_DIR" -- -j${PROC}

View File

@@ -10,14 +10,26 @@ set_component_src hipSPARSE
build_hipsparse() {
echo "Start build"
CXX="g++"
CXX_FLAG=
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
CXX="${ROCM_PATH}/llvm/bin/clang++"
CXX_FLAG="-DCMAKE_CXX_COMPILER=${ROCM_PATH}/llvm/bin/clang++"
fi
cd $COMPONENT_SRC
CXX="g++"
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
set_address_sanitizer_on
fi
SHARED_LIBS="ON"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
SHARED_LIBS="OFF"
fi
echo "C compiler: $CC"
echo "CXX compiler: $CXX"
@@ -25,15 +37,16 @@ build_hipsparse() {
init_rocm_common_cmake_params
cmake \
-DCPACK_SET_DESTDIR=OFF \
${LAUNCHER_FLAGS} \
"${rocm_math_common_cmake_params[@]}" \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
-DUSE_CUDA=OFF \
-DBUILD_CLIENTS_SAMPLES=ON \
-DBUILD_CLIENTS_TESTS=ON \
-DCMAKE_INSTALL_PREFIX=${ROCM_PATH} \
-DCMAKE_MODULE_PATH="${ROCM_PATH}/lib/cmake/hip;${ROCM_PATH}/hip/cmake" \
-DBUILD_ADDRESS_SANITIZER="${ADDRESS_SANITIZER}" \
${CXX_FLAG} \
"$COMPONENT_SRC"
cmake --build "$BUILD_DIR" -- -j${PROC}

View File

@@ -21,6 +21,10 @@ done
build_hipsparselt() {
echo "Start build"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
ack_and_skip_static
fi
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
set_address_sanitizer_on
@@ -50,7 +54,6 @@ build_hipsparselt() {
-DBUILD_CLIENTS_SAMPLES=ON \
-DBUILD_CLIENTS_TESTS=ON \
-DBUILD_CLIENTS_BENCHMARKS=ON \
-DCPACK_SET_DESTDIR=OFF \
-DCMAKE_INSTALL_PREFIX=${ROCM_PATH} \
-DBUILD_ADDRESS_SANITIZER="${ADDRESS_SANITIZER}" \
"$COMPONENT_SRC"

View File

@@ -9,6 +9,10 @@ set_component_src hipTensor
build_hiptensor() {
echo "Start build hipTensor"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
ack_and_skip_static
fi
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
set_address_sanitizer_on
@@ -18,7 +22,6 @@ build_hiptensor() {
mkdir -p "$BUILD_DIR" && cd "$BUILD_DIR"
init_rocm_common_cmake_params
if [ -n "$GPU_ARCHS" ]; then
GPU_TARGETS="$GPU_ARCHS"
else

View File

@@ -1,135 +0,0 @@
#!/bin/bash
source "$(dirname "${BASH_SOURCE}")/compute_utils.sh"
printUsage() {
echo
echo "Usage: $(basename "${BASH_SOURCE}") [options ...] [make options]"
echo
echo "Options:"
echo " -c, --clean Clean output and delete all intermediate work"
echo " -r, --release Make a release build instead of a debug build"
echo " -a, --address_sanitizer Enable address sanitizer"
echo " -o, --outdir <pkg_type> Print path of output directory containing packages of type referred to by pkg_type"
echo " -h, --help Prints this help"
echo " -s, --static Build static lib (.a). build instead of dynamic/shared(.so) "
echo
echo
return 0
}
TARGET="build"
PACKAGE_ROOT="$(getPackageRoot)"
PACKAGE_SRC="$(getSrcPath)"
PACKAGE_LIB="$(getLibPath)"
PACKAGE_BIN="$(getBinPath)"
PACKAGE_DEB="$(getPackageRoot)/deb/rocr"
PACKAGE_RPM="$(getPackageRoot)/rpm/rocr"
MAKEARG=""
CORE_BUILD_DIR="$(getBuildPath hsa-core)"
ROCR_DEV_BUILD_DIR="$(getBuildPath hsa-rocr-dev)"
PREFIX_PATH="$PACKAGE_ROOT"
BUILD_TYPE="Debug"
SHARED_LIBS="ON"
CLEAN_OR_OUT=0;
MAKETARGET="deb"
PKGTYPE="deb"
unset HIP_DEVICE_LIB_PATH
unset ROCM_PATH
VALID_STR=`getopt -o hcraso: --long help,clean,release,static,address_sanitizer,outdir: -- "$@"`
eval set -- "$VALID_STR"
while true ;
do
case "$1" in
(-h | --help)
printUsage ; exit 0;;
(-c | --clean)
TARGET="clean" ; ((CLEAN_OR_OUT|=1)) ; shift ;;
(-r | --release)
BUILD_TYPE="RelWithDebInfo" ; shift ;;
(-a | --address_sanitizer)
set_asan_env_vars
set_address_sanitizer_on ; shift ;;
(-s | --static)
SHARED_LIBS="OFF" ; shift ;;
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 2 ;;
--) shift; break;;
(*)
echo " This should never come but just incase : UNEXPECTED ERROR Parm : [$1] ">&2 ; exit 20;;
esac
done
RET_CONFLICT=1
check_conflicting_options $CLEAN_OR_OUT $PKGTYPE $MAKETARGET
if [ $RET_CONFLICT -ge 30 ]; then
print_vars $API_NAME $TARGET $BUILD_TYPE $SHARED_LIBS $CLEAN_OR_OUT $PKGTYPE $MAKETARGET
exit $RET_CONFLICT
fi
clean_hsa() {
echo "Cleaning HSA"
rm -rf "$CORE_BUILD_DIR"
rm -rf "$PACKAGE_RPM"
rm -rf "$PACKAGE_DEB"
rm -f "$PACKAGE_ROOT"/lib/libhsa-runtime*
rm -rf "$PACKAGE_ROOT/lib/cmake/hsa-runtime64"
rm -rf "$PACKAGE_ROOT/include/hsa"
rm -rf "$PACKAGE_ROOT/share/doc/hsa-runtime64"
rm -rf "$PACKAGE_ROOT/hsa"
}
build_hsa_core() {
echo "Build HSA"
local coreMakeOpts="$DASH_JAY -C $CORE_BUILD_DIR"
echo "$HSA_CORE_ROOT"
if [ ! -d "$CORE_BUILD_DIR" ]; then
mkdir -p "$CORE_BUILD_DIR"
pushd "$CORE_BUILD_DIR"
print_lib_type $SHARED_LIBS
cmake $(rocm_cmake_params) \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
-DENABLE_LDCONFIG=OFF \
$(rocm_common_cmake_params) \
-DADDRESS_SANITIZER="$ADDRESS_SANITIZER" \
"$HSA_CORE_ROOT"
popd
fi
time cmake --build "$CORE_BUILD_DIR" -- $coreMakeOpts
time cmake --build "$CORE_BUILD_DIR" -- $coreMakeOpts install
time cmake --build "$CORE_BUILD_DIR" -- $coreMakeOpts package
copy_if DEB "${CPACKGEN:-"DEB;RPM"}" "$PACKAGE_DEB" $CORE_BUILD_DIR/hsa-rocr*.deb
copy_if RPM "${CPACKGEN:-"DEB;RPM"}" "$PACKAGE_RPM" $CORE_BUILD_DIR/hsa-rocr*.rpm
}
print_output_directory() {
case ${PKGTYPE} in
("deb")
echo ${PACKAGE_DEB};;
("rpm")
echo ${PACKAGE_RPM};;
(*)
echo "Invalid package type \"${PKGTYPE}\" provided for -o" >&2; exit 1;;
esac
exit
}
case $TARGET in
(clean) clean_hsa ;;
(build) build_hsa_core;;
(outdir) print_output_directory ;;
(*) die "Invalid target $TARGET" ;;
esac
echo "Operation complete"

View File

@@ -11,7 +11,6 @@ printUsage() {
echo "Usage: $(basename "${BASH_SOURCE}") [options ...]"
echo
echo "Options:"
echo " -t, --alt Build the 'alt' variant"
echo " -c, --clean Clean output and delete all intermediate work"
echo " -d, --debug Build a debug version of llvm (excludes packaging)"
echo " -r, --release Build a release version of the package"
@@ -33,35 +32,26 @@ printUsage() {
return 0
}
PROJ_NAME="lightning"
ROCM_LLVM_LIB_RPATH='\$ORIGIN'
ROCM_LLVM_EXE_RPATH='\$ORIGIN/../lib:\$ORIGIN/../../../lib'
PACKAGE_OUT="$(getPackageRoot)"
BUILD_PATH="$(getBuildPath lightning)"
DEB_PATH="$(getDebPath lightning)"
RPM_PATH="$(getRpmPath lightning)"
BUILD_PATH="$(getBuildPath $PROJ_NAME)"
DEB_PATH="$(getDebPath $PROJ_NAME)"
RPM_PATH="$(getRpmPath $PROJ_NAME)"
INSTALL_PATH="${ROCM_INSTALL_PATH}/lib/llvm"
LLVM_ROOT_LCL="${LLVM_ROOT}"
ROCM_WHEEL_DIR="${BUILD_PATH}/_wheel"
TARGET="all"
MAKEOPTS="$DASH_JAY"
BUILD_TYPE="Release"
case "${JOB_NAME}" in
( *"rel"* | \
*"afar"* | \
*"nfar"* )
ENABLE_ASSERTIONS=0 ;;
( * )
ENABLE_ASSERTIONS=1 ;;
esac
ENABLE_ASSERTIONS=0
SHARED_LIBS="ON"
BUILD_LLVM_DYLIB="OFF"
FLANG_NEW=0
BUILD_ALT=0
CLEAN_OR_OUT=0;
PKGTYPE="deb"
MAKETARGET="deb"
@@ -74,10 +64,10 @@ BUILD_MANPAGES="ON"
STATIC_FLAG=
SANITIZER_AMDGPU=1
HSA_INC_PATH="$WORK_ROOT/ROCR-Runtime/src/inc"
COMGR_INC_PATH="$WORK_ROOT/llvm-project/amd/comgr/include"
HSA_INC_PATH="$WORK_ROOT/ROCR-Runtime/runtime/hsa-runtime/inc/"
COMGR_INC_PATH="$COMGR_ROOT/include"
VALID_STR=`getopt -o htcV:v:draAswlo:BPNM --long help,alt,clean,assert_llvm_ver_major:,assert_llvm_ver_minor:,debug,release,address_sanitizer,no_address_sanitizer,static,build_llvm_static,wheel,build,package,skip_lit_tests,skip_man_pages,outdir: -- "$@"`
VALID_STR=`getopt -o hcV:v:draAswlo:BPNM --long help,clean,assert_llvm_ver_major:,assert_llvm_ver_minor:,debug,release,address_sanitizer,no_address_sanitizer,static,build_llvm_static,wheel,build,package,skip_lit_tests,skip_man_pages,outdir: -- "$@"`
eval set -- "$VALID_STR"
set_dwarf_version(){
@@ -96,11 +86,10 @@ set_dwarf_version(){
while true ;
do
#echo "processing $1"
case "$1" in
(-h | --help)
printUsage ; exit 0;;
(-t | --alt)
BUILD_ALT=1 ; shift ;;
(-c | --clean)
TARGET="clean" ; ((CLEAN_OR_OUT|=1)) ; shift ;;
(-V | --assert_llvm_ver_major)
@@ -115,7 +104,7 @@ do
set_dwarf_version
SANITIZER_AMDGPU=1 ;
HSA_INC_PATH="$WORK_ROOT/hsa/runtime/opensrc/hsa-runtime/inc" ;
COMGR_INC_PATH="$WORK_ROOT/external/llvm-project/amd/comgr/include" ; shift ;;
COMGR_INC_PATH="$COMGR_ROOT/include" ; shift ;;
(-A | --no_address_sanitizer)
SANITIZER_AMDGPU=0 ;
unset HSA_INC_PATH ;
@@ -155,24 +144,11 @@ LLVM_PROJECTS="clang;lld;clang-tools-extra"
ENABLE_RUNTIMES="compiler-rt;libunwind"
BOOTSTRAPPING_BUILD_LIBCXX=0
BUILD_AMDCLANG="ON"
if [ $BUILD_ALT -eq 1 ]; then
BUILD_PATH="${BUILD_PATH}-alt"
DEB_PATH="${DEB_PATH}-alt"
RPM_PATH="${RPM_PATH}-alt"
INSTALL_PATH="${INSTALL_PATH}/alt"
LLVM_ROOT_LCL="${LLVM_ALT_ROOT}"
BUILD_AMDCLANG="OFF"
BUILD_MANPAGES="OFF"
SANITIZER_AMDGPU=0
unset HSA_INC_PATH
unset COMGR_INC_PATH
else
ENABLE_RUNTIMES="$ENABLE_RUNTIMES;libcxx;libcxxabi";
BOOTSTRAPPING_BUILD_LIBCXX=1
fi
ENABLE_RUNTIMES="$ENABLE_RUNTIMES;libcxx;libcxxabi"
BOOTSTRAPPING_BUILD_LIBCXX=1
clean_lightning() {
rm -rf "$ROCM_WHEEL_DIR"
rm -rf "$BUILD_PATH"
rm -rf "$DEB_PATH"
rm -rf "$RPM_PATH"
@@ -188,22 +164,14 @@ setup_llvm_info() {
local LLVM_URL_BRANCH
if [[ "${JOB_NAME}" == *rel* ]]; then
if [ $BUILD_ALT -eq 1 ]; then
LLVM_URL_BRANCH=$(git rev-parse HEAD)
else
LLVM_URL_NAME="https://github.com/RadeonOpenCompute/llvm-project"
LLVM_BRANCH_NAME="roc-${ROCM_VERSION}"
LLVM_URL_BRANCH="${LLVM_URL_NAME} ${LLVM_BRANCH_NAME}"
fi
else
LLVM_REMOTE_NAME=$(git remote)
LLVM_URL_NAME=$(git config --get remote."${LLVM_REMOTE_NAME}".url)
if [ $BUILD_ALT -eq 1 ]; then
LLVM_BRANCH_NAME=$(repo manifest | sed -n 's/.*path="external\/llvm-project-alt\/llvm-project".* upstream="\([^"]*\)".*/\1/p' )
else
LLVM_REMOTE_NAME=$(git remote)
LLVM_URL_NAME=$(git config --get remote."${LLVM_REMOTE_NAME}".url)
LLVM_BRANCH_NAME=$(repo manifest | sed -n 's/.*path="external\/llvm-project".* upstream="\([^"]*\)".*/\1/p' )
fi
LLVM_URL_BRANCH="${LLVM_URL_NAME} ${LLVM_BRANCH_NAME}"
LLVM_URL_BRANCH="${LLVM_URL_NAME} ${LLVM_BRANCH_NAME}"
fi
LLVM_COMMIT_GITDATE=$(git show -s --format=@%ct | xargs | date -f - --utc +%y%U%w)
@@ -283,24 +251,27 @@ build_lightning() {
mkdir -p "$BUILD_PATH"
pushd "$BUILD_PATH"
eval EXTRA_LLVM_CMAKE_PARAMS_ARRAY=($EXTRA_LLVM_CMAKE_PARAMS)
if [ ! -e Makefile ]; then
echo "Building LLVM CMake environment"
if [ -e "$LLVM_ROOT_LCL/../flang/AFARrelease" ]; then
LLVM_PROJECTS="$LLVM_PROJECTS;mlir"
if [ -e "$LLVM_ROOT_LCL/../flang/EnableFlangBuild" ]; then
FLANG_NEW=1
LLVM_PROJECTS="$LLVM_PROJECTS;flang;mlir"
LLVM_PROJECTS="$LLVM_PROJECTS;flang"
ENABLE_RUNTIMES="$ENABLE_RUNTIMES;openmp";
else
if [[ "${JOB_NAME}" != *afar* ]] && [ -e "$LLVM_ROOT_LCL/../flang/DoROCmRelease" ]; then
FLANG_NEW=1
LLVM_PROJECTS="$LLVM_PROJECTS;flang;mlir"
else
echo "NOT building project flang"
fi
if [[ "${JOB_NAME}" != *afar* ]] && [ -e "$LLVM_ROOT_LCL/../flang/DoROCmRelease" ]; then
FLANG_NEW=1
LLVM_PROJECTS="$LLVM_PROJECTS;flang"
else
echo "NOT building project flang"
fi
fi
set -x
cmake $(rocm_cmake_params) ${GEN_NINJA} \
${STATIC_FLAG} \
${PYTHON_VERSION_WORKAROUND} \
-DCMAKE_INSTALL_PREFIX="$INSTALL_PATH" \
-DLLVM_TARGETS_TO_BUILD="AMDGPU;X86" \
-DLLVM_ENABLE_PROJECTS="$LLVM_PROJECTS" \
@@ -342,6 +313,7 @@ build_lightning() {
-DCLANG_LINK_FLANG_LEGACY=ON \
-DCMAKE_CXX_STANDARD=17 \
-DFLANG_INCLUDE_DOCS=OFF \
"${EXTRA_LLVM_CMAKE_PARAMS_ARRAY[@]}" \
"$LLVM_ROOT_LCL"
set +x
echo "CMake complete"
@@ -358,28 +330,11 @@ build_lightning() {
echo "End Workaround for race condition"
cmake --build . -- $MAKEOPTS
case "$DISTRO_ID" in
(rhel*|centos*)
RHEL_BUILD=1
;;
(*)
RHEL_BUILD=0
;;
esac
if [ $SKIP_LIT_TESTS -eq 0 ]; then
if [ $RHEL_BUILD -eq 1 ] && [ $BUILD_ALT != 1 ]; then
if [ $FLANG_NEW -eq 1 ]; then
cmake --build . -- $MAKEOPTS check-lld check-mlir
else
cmake --build . -- $MAKEOPTS check-lld
fi
elif [ "$DISTRO_NAME" != "sles" ] && [ $BUILD_ALT != 1 ]; then
if [ $FLANG_NEW -eq 1 ]; then
cmake --build . -- $MAKEOPTS check-llvm check-clang check-lld check-mlir
else
cmake --build . -- $MAKEOPTS check-llvm check-clang check-lld
fi
if [ $RHEL_BUILD -eq 1 ]; then
cmake --build . -- $MAKEOPTS check-lld check-mlir
elif [ "$DISTRO_NAME" != "sles" ]; then
cmake --build . -- $MAKEOPTS check-llvm check-clang check-lld check-mlir
fi
fi
cmake --build . -- $MAKEOPTS clang-tidy
@@ -396,23 +351,15 @@ package_lightning_dynamic(){
get_llvm_version
local llvmParsedVersion="${LLVM_VERSION_MAJOR}.${LLVM_VERSION_MINOR}.${LLVM_VERSION_PATCH}"
local packageName="rocm-llvm"
local packageSummary="ROCm compiler"
local packageSummaryLong="ROCm compiler based on LLVM $llvmParsedVersion"
local installPath="$ROCM_INSTALL_PATH/lib/llvm/"
if [ $BUILD_ALT -eq 1 ]; then
local packageName="rocm-llvm-alt"
local packageSummary="Proprietary ROCm compiler"
local packageSummaryLong="ROCm compiler, including proprietary optimizations, based on LLVM $llvmParsedVersion"
local installPath="$ROCM_INSTALL_PATH/lib/llvm/alt"
else
local packageName="rocm-llvm"
local packageSummary="ROCm compiler"
local packageSummaryLong="ROCm compiler based on LLVM $llvmParsedVersion"
local installPath="$ROCM_INSTALL_PATH/lib/llvm/"
if [ "$BUILD_LLVM_DYLIB" == "ON" ] ; then
local packageNameCore="rocm-llvm-core"
local packageSummaryCore="ROCm core compiler dylibs"
local packageSummaryLongCore="ROCm compiler based on LLVM $llvmParsedVersion"
fi
if [ "$BUILD_LLVM_DYLIB" == "ON" ] ; then
local packageNameCore="rocm-llvm-core"
local packageSummaryCore="ROCm core compiler dylibs"
local packageSummaryLongCore="ROCm compiler based on LLVM $llvmParsedVersion"
fi
local packageArch="amd64"
@@ -433,9 +380,6 @@ package_lightning_dynamic(){
local prermFile="$packageDeb/DEBIAN/prerm"
local specFile="$packageDir/$packageName.spec"
local debDependencies="python3, libc6, libstdc++6|libstdc++8, libstdc++-5-dev|libstdc++-7-dev|libstdc++-11-dev, libgcc-5-dev|libgcc-7-dev|libgcc-11-dev, rocm-core"
if [ $BUILD_ALT -eq 1 ]; then
debDependencies="${debDependencies}, rocm-llvm"
fi
local debRecommends="gcc, g++, gcc-multilib, g++-multilib"
local packageRpm="$packageDir/rpm"
@@ -508,42 +452,33 @@ package_lightning_dynamic(){
debDependencies="${debDependencies}, ${packageNameCore}"
fi
if [ $BUILD_ALT -eq 0 ] ; then
cp -r "$LLVM_ROOT_LCL/LICENSE.TXT" "$packageDeb/$licenseDir"
else
cp -r "$LLVM_PROJECT_ALT_ROOT/EULA" "$packageDeb/$licenseDir"
cp -r "$LLVM_PROJECT_ALT_ROOT/DISCLAIMER.txt" "$packageDeb/$licenseDir"
fi
cp -r "$LLVM_ROOT_LCL/LICENSE.TXT" "$packageDeb/$licenseDir"
cp -r "$distBin" "$packageDeb/$installPath/bin"
cp -r "$distInc" "$packageDeb/$installPath/include"
cp -r "$distLib" "$packageDeb/$installPath/lib"
if [ "$BUILD_MANPAGES" == "ON" ]; then
if [ $BUILD_ALT -eq 0 ]; then
for i in "${man_pages[@]}"; do
gzip -f "$distMan/man1/$i"
for i in "${man_pages[@]}"; do
gzip -f "$distMan/man1/$i"
done
if [ -f "$distMan/man1/clang.1.gz" ]; then
for i in "${amd_man_pages[@]}"; do
ln -sf "clang.1.gz" "$distMan/man1/$i"
done
if [ -f "$distMan/man1/clang.1.gz" ]; then
for i in "${amd_man_pages[@]}"; do
ln -sf "clang.1.gz" "$distMan/man1/$i"
done
fi
fi
fi
cp -r "$distMan" "$packageDeb/$installPath/share"
if [ $BUILD_ALT -eq 0 ]; then
touch "$postinstFile" "$prermFile"
echo "mkdir -p \"$ROCM_INSTALL_PATH/bin\"" >> $postinstFile
for i in "${amd_compiler_commands[@]}"; do
if [ -f "$packageDeb/$installPath/bin/$i" ]; then
echo "ln -s \"../lib/llvm/bin/$i\" \"$ROCM_INSTALL_PATH/bin/$i\"" >> $postinstFile
echo "rm -f \"$ROCM_INSTALL_PATH/bin/$i\"" >> $prermFile
fi
done
echo "rmdir --ignore-fail-on-non-empty \"$ROCM_INSTALL_PATH/bin\"" >> $prermFile
chmod 0555 "$postinstFile" "$prermFile"
cp -P "$backwardsCompatibleSymlink" "$packageDeb/$ROCM_INSTALL_PATH"
fi
touch "$postinstFile" "$prermFile"
echo "mkdir -p \"$ROCM_INSTALL_PATH/bin\"" >> $postinstFile
for i in "${amd_compiler_commands[@]}"; do
if [ -f "$packageDeb/$installPath/bin/$i" ]; then
echo "ln -s \"../lib/llvm/bin/$i\" \"$ROCM_INSTALL_PATH/bin/$i\"" >> $postinstFile
echo "rm -f \"$ROCM_INSTALL_PATH/bin/$i\"" >> $prermFile
fi
done
echo "rmdir --ignore-fail-on-non-empty \"$ROCM_INSTALL_PATH/bin\"" >> $prermFile
chmod 0555 "$postinstFile" "$prermFile"
cp -P "$backwardsCompatibleSymlink" "$packageDeb/$ROCM_INSTALL_PATH"
echo "Package: $packageName" > $controlFile
echo "Architecture: $packageArch" >> $controlFile
@@ -613,16 +548,12 @@ package_lightning_dynamic(){
echo "Release: ${JOB_DESIGNATOR}${SLES_BUILD_ID_PREFIX}${BUILD_ID}%{?dist}" >> $specFile
echo "Summary: $packageSummary" >> $specFile
echo "Group: System Environment/Libraries" >> $specFile
if [ $BUILD_ALT -eq 1 ]; then
echo "License: AMD Proprietary" >> $specFile
else
echo "License: ASL 2.0 with exceptions" >> $specFile
fi
echo "License: ASL 2.0 with exceptions" >> $specFile
echo "Requires: $rpmRequires" >> $specFile
if [ $BUILD_ALT -eq 1 ]; then
echo "%define _build_id_links none" >> $specFile
fi
# The following is commented as Centos 7 has a version of rpm
# that does not understand it. When we no longer support Centos 7
# then we should have a correct recommends line.
#echo "Recommends: $rpmRecommends" >> $specFile
echo "%description" >> $specFile
echo "$packageSummaryLong" >> $specFile
@@ -638,28 +569,20 @@ package_lightning_dynamic(){
echo "mkdir -p \$RPM_BUILD_ROOT/$installPath/share/man" >> $specFile
echo "mkdir -p \$RPM_BUILD_ROOT/$licenseDir" >> $specFile
if [ $BUILD_ALT -eq 0 ]; then
echo "cp -R $LLVM_ROOT_LCL/LICENSE.TXT \$RPM_BUILD_ROOT/$licenseDir" >> $specFile
echo "cp -P $backwardsCompatibleSymlink \$RPM_BUILD_ROOT/$ROCM_INSTALL_PATH" >> $specFile
else
echo "cp -R $LLVM_PROJECT_ALT_ROOT/EULA \$RPM_BUILD_ROOT/$licenseDir" >> $specFile
echo "cp -R $LLVM_PROJECT_ALT_ROOT/DISCLAIMER.txt \$RPM_BUILD_ROOT/$licenseDir" >> $specFile
fi
echo "cp -R $LLVM_ROOT_LCL/LICENSE.TXT \$RPM_BUILD_ROOT/$licenseDir" >> $specFile
echo "cp -P $backwardsCompatibleSymlink \$RPM_BUILD_ROOT/$ROCM_INSTALL_PATH" >> $specFile
echo "cp -R $distBin \$RPM_BUILD_ROOT/$installPath" >> $specFile
echo "cp -R $distInc \$RPM_BUILD_ROOT/$installPath" >> $specFile
echo "cp -R $distLib \$RPM_BUILD_ROOT/$installPath" >> $specFile
if [ "$BUILD_MANPAGES" == "ON" ]; then
if [ $BUILD_ALT -eq 0 ]; then
for i in "${man_pages[@]}"; do
echo "gzip -f $distMan/man1/$i" >> $specFile
done
if [ -f "$distMan/man1/clang.1.gz" ]; then
for i in "${amd_man_pages[@]}"; do
echo "ln -sf clang.1.gz \"$distMan/man1/$i\"" >> $specFile
done
fi
fi
for i in "${man_pages[@]}"; do
echo "gzip -f $distMan/man1/$i" >> $specFile
done
if [ -f "$distMan/man1/clang.1.gz" ]; then
for i in "${amd_man_pages[@]}"; do
echo "ln -sf clang.1.gz \"$distMan/man1/$i\"" >> $specFile
done
fi
fi
echo "cp -R $distMan \$RPM_BUILD_ROOT/$installPath/share" >> $specFile
@@ -676,25 +599,20 @@ package_lightning_dynamic(){
echo "$ROCM_INSTALL_PATH" >> $specFile
echo "%post" >> $specFile
if [ $BUILD_ALT -eq 0 ]; then
echo "mkdir -p \"$ROCM_INSTALL_PATH/bin\"" >> $specFile
for i in "${amd_compiler_commands[@]}"; do
if [ -f "$distBin/$i" ]; then
echo "ln -sf ../lib/llvm/bin/$i \"$ROCM_INSTALL_PATH/bin/$i\"" >> $specFile
fi
done
fi
echo "mkdir -p \"$ROCM_INSTALL_PATH/bin\"" >> $specFile
for i in "${amd_compiler_commands[@]}"; do
if [ -f "$distBin/$i" ]; then
echo "ln -sf ../lib/llvm/bin/$i \"$ROCM_INSTALL_PATH/bin/$i\"" >> $specFile
fi
done
echo "%preun" >> $specFile
if [ $BUILD_ALT -eq 0 ]; then
for i in "${amd_compiler_commands[@]}"; do
if [ -f "$distBin/$i" ]; then
echo "rm -f \"$ROCM_INSTALL_PATH/bin/$i\"" >> $specFile
fi
done
echo "rmdir --ignore-fail-on-non-empty \"$ROCM_INSTALL_PATH/bin\"" >> $specFile
fi
for i in "${amd_compiler_commands[@]}"; do
if [ -f "$distBin/$i" ]; then
echo "rm -f \"$ROCM_INSTALL_PATH/bin/$i\"" >> $specFile
fi
done
echo "rmdir --ignore-fail-on-non-empty \"$ROCM_INSTALL_PATH/bin\"" >> $specFile
echo "%postun" >> $specFile
rpmbuild --define "_topdir $packageRpm" -ba $specFile
@@ -711,32 +629,17 @@ package_lightning_static() {
get_llvm_version
local llvmParsedVersion="${LLVM_VERSION_MAJOR}.${LLVM_VERSION_MINOR}.${LLVM_VERSION_PATCH}"
if [ $BUILD_ALT -eq 1 ]; then
local packageName="rocm-llvm-alt"
local packageSummary="Proprietary ROCm core compiler"
local packageSummaryLong="ROCm core compiler, including proprietary optimizations based on LLVM $llvmParsedVersion"
if [ "$PACKAGEEXT" = "deb" ]; then
local packageNameExtra="rocm-llvm-alt-dev"
else
local packageNameExtra="rocm-llvm-alt-devel"
fi
local packageSummaryExtra="Proprietary ROCm compiler dev tools"
local packageSummaryLongExtra="ROCm compiler dev tools and documentation, including proprietary optimizations, based on LLVM $llvmParsedVersion"
local installPath="$ROCM_INSTALL_PATH/lib/llvm/alt"
local packageName="rocm-llvm"
local packageSummary="ROCm core compiler"
local packageSummaryLong="ROCm core compiler based on LLVM $llvmParsedVersion"
if [ "$PACKAGEEXT" = "deb" ]; then
local packageNameExtra="rocm-llvm-dev"
else
local packageName="rocm-llvm"
local packageSummary="ROCm core compiler"
local packageSummaryLong="ROCm core compiler based on LLVM $llvmParsedVersion"
if [ "$PACKAGEEXT" = "deb" ]; then
local packageNameExtra="rocm-llvm-dev"
else
local packageNameExtra="rocm-llvm-devel"
fi
local packageSummaryExtra="ROCm compiler dev tools"
local packageSummaryLongExtra="ROCm compiler dev tools and documentation, based on LLVM $llvmParsedVersion"
local installPath="$ROCM_INSTALL_PATH/lib/llvm/"
local packageNameExtra="rocm-llvm-devel"
fi
local packageSummaryExtra="ROCm compiler dev tools"
local packageSummaryLongExtra="ROCm compiler dev tools and documentation, based on LLVM $llvmParsedVersion"
local installPath="$ROCM_INSTALL_PATH/lib/llvm/"
local packageArch="amd64"
local packageVersion="${llvmParsedVersion}.${LLVM_COMMIT_GITDATE}"
@@ -746,7 +649,6 @@ package_lightning_static() {
local distLib="$INSTALL_PATH/lib"
local distMan="$INSTALL_PATH/share/man"
local licenseDir="$ROCM_INSTALL_PATH/share/doc/$packageName"
local licenseDirExtra="$ROCM_INSTALL_PATH/share/doc/$packageNameExtra"
local packageDir="$BUILD_PATH/package"
local backwardsCompatibleSymlink="$ROCM_INSTALL_PATH/llvm"
@@ -756,9 +658,6 @@ package_lightning_static() {
local prermFile="$packageDeb/DEBIAN/prerm"
local specFile="$packageDir/$packageName.spec"
local debDependencies="python3, libc6, libstdc++6|libstdc++8, libstdc++-5-dev|libstdc++-7-dev|libstdc++-11-dev, libgcc-5-dev|libgcc-7-dev|libgcc-11-dev, rocm-core"
if [ $BUILD_ALT -eq 1 ]; then
debDependencies="${debDependencies}, rocm-llvm"
fi
local debRecommends="gcc, g++, gcc-multilib, g++-multilib"
local packageRpm="$packageDir/rpm"
@@ -767,10 +666,6 @@ package_lightning_static() {
local specFileExtra="$packageDir/$packageNameExtra.spec"
local rpmRequires="rocm-core"
local rpmRequiresExtra="rocm-core, $packageName"
if [ $BUILD_ALT -eq 1 ]; then
rpmRequires+=", rocm-llvm"
rpmRequiresExtra+=", rocm-llvm-devel"
fi
local rpmRecommends="gcc, gcc-c++, devtoolset-7-gcc-c++"
rm -rf "$packageDir"
@@ -807,12 +702,7 @@ package_lightning_static() {
mkdir -p "$DEB_PATH"
mkdir -p "$packageDeb/$licenseDir"
if [ $BUILD_ALT -eq 0 ] ; then
cp -r "$LLVM_ROOT_LCL/LICENSE.TXT" "$packageDeb/$licenseDir"
else
cp -r "$LLVM_PROJECT_ALT_ROOT/EULA" "$packageDeb/$licenseDir"
cp -r "$LLVM_PROJECT_ALT_ROOT/DISCLAIMER.txt" "$packageDeb/$licenseDir"
fi
cp -r "$LLVM_ROOT_LCL/LICENSE.TXT" "$packageDeb/$licenseDir"
mkdir -p "$packageDeb/$installPath/bin"
for i in "${core_bin[@]}"; do
@@ -838,36 +728,32 @@ package_lightning_static() {
done
if [ "$BUILD_MANPAGES" == "ON" ]; then
if [ $BUILD_ALT -eq 0 ]; then
mkdir -p "$packageDeb/$installPath/share/man1"
for i in "${core_man_pages[@]}"; do
if [ -f "$distMan/man1/$i" ]; then
gzip -f "$distMan/man1/$i"
cp -d "$distMan/man1/${i}.gz" "$packageDeb/$installPath/share/man1/"
fi
done
if [ -f "$distMan/man1/clang.1.gz" ]; then
for i in "${amd_man_pages[@]}"; do
ln -sf "clang.1.gz" "$distMan/man1/$i"
cp -d "$distMan/man1/${i}" "$packageDeb/$installPath/share/man1/"
done
mkdir -p "$packageDeb/$installPath/share/man1"
for i in "${core_man_pages[@]}"; do
if [ -f "$distMan/man1/$i" ]; then
gzip -f "$distMan/man1/$i"
cp -d "$distMan/man1/${i}.gz" "$packageDeb/$installPath/share/man1/"
fi
done
if [ -f "$distMan/man1/clang.1.gz" ]; then
for i in "${amd_man_pages[@]}"; do
ln -sf "clang.1.gz" "$distMan/man1/$i"
cp -d "$distMan/man1/${i}" "$packageDeb/$installPath/share/man1/"
done
fi
fi
if [ $BUILD_ALT -eq 0 ]; then
touch "$postinstFile" "$prermFile"
echo "mkdir -p \"$ROCM_INSTALL_PATH/bin\"" >> $postinstFile
for i in "${amd_compiler_commands[@]}"; do
if [ -f "$packageDeb/$installPath/bin/$i" ]; then
echo "ln -s \"../lib/llvm/bin/$i\" \"$ROCM_INSTALL_PATH/bin/$i\"" >> $postinstFile
echo "rm -f \"$ROCM_INSTALL_PATH/bin/$i\"" >> $prermFile
fi
done
echo "rmdir --ignore-fail-on-non-empty \"$ROCM_INSTALL_PATH/bin\"" >> $prermFile
chmod 0555 "$postinstFile" "$prermFile"
cp -P "$backwardsCompatibleSymlink" "$packageDeb/$ROCM_INSTALL_PATH"
fi
touch "$postinstFile" "$prermFile"
echo "mkdir -p \"$ROCM_INSTALL_PATH/bin\"" >> $postinstFile
for i in "${amd_compiler_commands[@]}"; do
if [ -f "$packageDeb/$installPath/bin/$i" ]; then
echo "ln -s \"../lib/llvm/bin/$i\" \"$ROCM_INSTALL_PATH/bin/$i\"" >> $postinstFile
echo "rm -f \"$ROCM_INSTALL_PATH/bin/$i\"" >> $prermFile
fi
done
echo "rmdir --ignore-fail-on-non-empty \"$ROCM_INSTALL_PATH/bin\"" >> $prermFile
chmod 0555 "$postinstFile" "$prermFile"
cp -P "$backwardsCompatibleSymlink" "$packageDeb/$ROCM_INSTALL_PATH"
{
echo "Package: $packageName"
@@ -892,14 +778,6 @@ package_lightning_static() {
mkdir -p "$packageDeb/$installPath"
mkdir "${controlFile%/*}"
mkdir -p "$DEB_PATH"
mkdir -p "$packageDeb/$licenseDirExtra"
if [ $BUILD_ALT -eq 0 ] ; then
cp -r "$LLVM_ROOT_LCL/LICENSE.TXT" "$packageDeb/$licenseDirExtra"
else
cp -r "$LLVM_PROJECT_ALT_ROOT/EULA" "$packageDeb/$licenseDirExtra"
cp -r "$LLVM_PROJECT_ALT_ROOT/DISCLAIMER.txt" "$packageDeb/$licenseDirExtra"
fi
mkdir -p "$packageDeb/$installPath/bin"
for i in "$distBin"/*; do
@@ -922,21 +800,16 @@ package_lightning_static() {
fi
if [ "$BUILD_MANPAGES" == "ON" ]; then
if [ $BUILD_ALT -eq 0 ]; then
mkdir -p "$packageDeb/$installPath/share/man1"
for i in "${dev_man_pages[@]}"; do
if [ -f "$distMan/man1/$i" ]; then
gzip -f "$distMan/man1/$i"
cp -d "$distMan/man1/${i}.gz" "$packageDeb/$installPath/share/man1/"
fi
done
fi
mkdir -p "$packageDeb/$installPath/share/man1"
for i in "${dev_man_pages[@]}"; do
if [ -f "$distMan/man1/$i" ]; then
gzip -f "$distMan/man1/$i"
cp -d "$distMan/man1/${i}.gz" "$packageDeb/$installPath/share/man1/"
fi
done
fi
debDependencies="${debDependencies}, ${packageName}"
if [ $BUILD_ALT -eq 1 ]; then
debDependencies="${debDependencies}, rocm-llvm-dev"
fi
echo "Package: $packageNameExtra" > $controlFile
echo "Architecture: $packageArch" >> $controlFile
@@ -979,13 +852,8 @@ package_lightning_static() {
echo "mkdir -p \$RPM_BUILD_ROOT/$installPath/bin" >> $specFile
echo "mkdir -p \$RPM_BUILD_ROOT/$licenseDir" >> $specFile
if [ $BUILD_ALT -eq 0 ]; then
echo "cp -R $LLVM_ROOT_LCL/LICENSE.TXT \$RPM_BUILD_ROOT/$licenseDir" >> $specFile
echo "cp -P $backwardsCompatibleSymlink \$RPM_BUILD_ROOT/$ROCM_INSTALL_PATH" >> $specFile
else
echo "cp -R $LLVM_PROJECT_ALT_ROOT/EULA \$RPM_BUILD_ROOT/$licenseDir" >> $specFile
echo "cp -R $LLVM_PROJECT_ALT_ROOT/DISCLAIMER.txt \$RPM_BUILD_ROOT/$licenseDir" >> $specFile
fi
echo "cp -R $LLVM_ROOT_LCL/LICENSE.TXT \$RPM_BUILD_ROOT/$licenseDir" >> $specFile
echo "cp -P $backwardsCompatibleSymlink \$RPM_BUILD_ROOT/$ROCM_INSTALL_PATH" >> $specFile
for i in "${core_bin[@]}"; do
if [ -f "$distBin/$i" ]; then
@@ -995,9 +863,7 @@ package_lightning_static() {
echo "cp -d \"$distBin/flang\" \$RPM_BUILD_ROOT/$installPath/bin/" >> $specFile
if [ $BUILD_ALT -eq 0 ]; then
echo "cp -d \"$distBin\"/*.cfg \$RPM_BUILD_ROOT/$installPath/bin/" >> $specFile
fi
echo "cp -d \"$distBin\"/*.cfg \$RPM_BUILD_ROOT/$installPath/bin/" >> $specFile
echo "mkdir -p \$RPM_BUILD_ROOT/$installPath/lib/clang" >> $specFile
echo "cp -R \"$distLib/clang/\" \$RPM_BUILD_ROOT/$installPath/lib/" >> $specFile
@@ -1014,20 +880,18 @@ package_lightning_static() {
done
if [ "$BUILD_MANPAGES" == "ON" ]; then
if [ $BUILD_ALT -eq 0 ]; then
echo "mkdir -p \$RPM_BUILD_ROOT/$installPath/share/man/man1" >> $specFile
for i in "${core_man_pages[@]}"; do
if [ -f "$distMan/man1/$i" ]; then
echo "gzip -f $distMan/man1/$i" >> $specFile
echo "cp -d $distMan/man1/${i}.gz \$RPM_BUILD_ROOT/$installPath/share/man/man1/" >> $specFile
fi
done
if [ -f "$distMan/man1/clang.1.gz" ]; then
for i in "${amd_man_pages[@]}"; do
echo "ln -sf clang.1.gz \"$distMan/man1/$i\"" >> $specFile
echo "cp -d $distMan/man1/${i} \$RPM_BUILD_ROOT/$installPath/share/man/man1/" >> $specFile
done
echo "mkdir -p \$RPM_BUILD_ROOT/$installPath/share/man/man1" >> $specFile
for i in "${core_man_pages[@]}"; do
if [ -f "$distMan/man1/$i" ]; then
echo "gzip -f $distMan/man1/$i" >> $specFile
echo "cp -d $distMan/man1/${i}.gz \$RPM_BUILD_ROOT/$installPath/share/man/man1/" >> $specFile
fi
done
if [ -f "$distMan/man1/clang.1.gz" ]; then
for i in "${amd_man_pages[@]}"; do
echo "ln -sf clang.1.gz \"$distMan/man1/$i\"" >> $specFile
echo "cp -d $distMan/man1/${i} \$RPM_BUILD_ROOT/$installPath/share/man/man1/" >> $specFile
done
fi
fi
@@ -1039,24 +903,20 @@ package_lightning_static() {
echo "$ROCM_INSTALL_PATH"
echo "%post"
if [ $BUILD_ALT -eq 0 ]; then
echo "mkdir -p \"$ROCM_INSTALL_PATH/bin\""
for i in "${amd_compiler_commands[@]}"; do
if [ -f "$distBin/$i" ]; then
echo "ln -sf ../lib/llvm/bin/$i \"$ROCM_INSTALL_PATH/bin/$i\""
fi
done
fi
echo "mkdir -p \"\$RPM_INSTALL_PREFIX0/bin\""
for i in "${amd_compiler_commands[@]}"; do
if [ -f "$distBin/$i" ]; then
echo "ln -sf ../lib/llvm/bin/$i \"\$RPM_INSTALL_PREFIX0/bin/$i\""
fi
done
echo "%preun"
if [ $BUILD_ALT -eq 0 ]; then
for i in "${amd_compiler_commands[@]}"; do
if [ -f "$distBin/$i" ]; then
echo "rm -f \"$ROCM_INSTALL_PATH/bin/$i\""
fi
done
echo "rmdir --ignore-fail-on-non-empty \"$ROCM_INSTALL_PATH/bin\""
fi
for i in "${amd_compiler_commands[@]}"; do
if [ -f "$distBin/$i" ]; then
echo "rm -f \"\$RPM_INSTALL_PREFIX0/bin/$i\""
fi
done
echo "rmdir --ignore-fail-on-non-empty \"\$RPM_INSTALL_PREFIX0/bin\""
echo "%postun"
} >> "$specFile"
@@ -1071,16 +931,13 @@ package_lightning_static() {
echo "Release: ${JOB_DESIGNATOR}${SLES_BUILD_ID_PREFIX}${BUILD_ID}%{?dist}" >> $specFileExtra
echo "Summary: $packageSummaryExtra" >> $specFileExtra
echo "Group: System Environment/Libraries" >> $specFileExtra
if [ $BUILD_ALT -eq 1 ]; then
echo "License: AMD Proprietary" >> $specFileExtra
else
echo "License: ASL 2.0 with exceptions" >> $specFileExtra
fi
echo "License: ASL 2.0 with exceptions" >> $specFileExtra
echo "Prefix: $ROCM_INSTALL_PATH" >> $specFileExtra
echo "Requires: $rpmRequiresExtra" >> $specFileExtra
if [ $BUILD_ALT -eq 1 ]; then
echo "%define _build_id_links none" >> $specFileExtra
fi
# The following is commented as Centos 7 has a version of rpm
# that does not understand it. When we no longer support Centos 7
# then we should have a correct recommends line.
#echo "Recommends: $rpmRecommends" >> $specFileExtra
echo "%description" >> $specFileExtra
echo "$packageSummaryLongExtra" >> $specFileExtra
@@ -1093,15 +950,8 @@ package_lightning_static() {
echo "mkdir -p \$RPM_BUILD_ROOT/$installPath/bin" >> $specFileExtra
echo "mkdir -p \$RPM_BUILD_ROOT/$installPath/include" >> $specFileExtra
echo "mkdir -p \$RPM_BUILD_ROOT/$installPath/lib" >> $specFileExtra
echo "mkdir -p \$RPM_BUILD_ROOT/$licenseDirExtra" >> $specFileExtra
if [ $BUILD_ALT -eq 0 ]; then
echo "cp -R $LLVM_ROOT_LCL/LICENSE.TXT \$RPM_BUILD_ROOT/$licenseDirExtra" >> $specFileExtra
echo "cp -P $backwardsCompatibleSymlink \$RPM_BUILD_ROOT/$ROCM_INSTALL_PATH" >> $specFileExtra
else
echo "cp -R $LLVM_PROJECT_ALT_ROOT/EULA \$RPM_BUILD_ROOT/$licenseDirExtra" >> $specFileExtra
echo "cp -R $LLVM_PROJECT_ALT_ROOT/DISCLAIMER.txt \$RPM_BUILD_ROOT/$licenseDirExtra" >> $specFileExtra
fi
echo "cp -P $backwardsCompatibleSymlink \$RPM_BUILD_ROOT/$ROCM_INSTALL_PATH" >> $specFileExtra
for i in "$distBin"/*; do
bin=$(basename "$i")
@@ -1122,15 +972,13 @@ package_lightning_static() {
fi
if [ "$BUILD_MANPAGES" == "ON" ]; then
if [ $BUILD_ALT -eq 0 ]; then
echo "mkdir -p \$RPM_BUILD_ROOT/$installPath/share/man/man1" >> $specFileExtra
for i in "${extra_man_pages[@]}"; do
if [ -f "$distMan/man1/$i" ]; then
echo "gzip -f $distMan/man1/$i" >> $specFileExtra
echo "cp -d \"$distMan/man1/${i}.gz\" \$RPM_BUILD_ROOT/$installPath/share/man/man1/" >> $specFileExtra
fi
done
fi
echo "mkdir -p \$RPM_BUILD_ROOT/$installPath/share/man/man1" >> $specFileExtra
for i in "${dev_man_pages[@]}"; do
if [ -f "$distMan/man1/$i" ]; then
echo "gzip -f $distMan/man1/$i" >> $specFileExtra
echo "cp -d \"$distMan/man1/${i}.gz\" \$RPM_BUILD_ROOT/$installPath/share/man/man1/" >> $specFileExtra
fi
done
fi
echo "%clean" >> $specFileExtra
@@ -1266,9 +1114,7 @@ print_output_directory() {
build() {
mkdir -p "${INSTALL_PATH}"
build_lightning
if [ $BUILD_ALT -eq 0 ] ; then
create_compiler_config_files
fi
create_compiler_config_files
}
create_wheel_package() {
@@ -1301,9 +1147,4 @@ case $TARGET in
(*) die "Invalid target $TARGET" ;;
esac
if [[ $WHEEL_PACKAGE == true ]]; then
echo "Wheel Package build started !!!!"
create_wheel_package
fi
echo "Operation complete"

View File

@@ -12,6 +12,10 @@ RPM_PATH=$PACKAGE_DIR
build_miopen_hip() {
echo "Start build"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
ack_and_skip_static
fi
cd $COMPONENT_SRC
git config --global --add safe.directory "$COMPONENT_SRC"
checkout_lfs
@@ -26,10 +30,12 @@ build_miopen_hip() {
cmake \
"${rocm_math_common_cmake_params[@]}" \
-DMIOPEN_BACKEND=HIP \
-DMIOPEN_OFFLINE_COMPILER_PATHS_V2=1 \
-DCMAKE_CXX_COMPILER="${ROCM_PATH}/llvm/bin/clang++" \
-DCMAKE_C_COMPILER="${ROCM_PATH}/llvm/bin/clang" \
-DCMAKE_PREFIX_PATH="${ROCM_PATH};${ROCM_PATH}/hip;${HOME}/miopen-deps" \
-DHIP_OC_COMPILER="${ROCM_PATH}/bin/clang-ocl" \
-DMIOPEN_TEST_DISCRETE=OFF \
"$COMPONENT_SRC"
cmake --build "$BUILD_DIR" -- -j${PROC}

View File

@@ -9,6 +9,10 @@ BUILD_DEV=ON
build_mivisionx() {
echo "Start build"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
ack_and_skip_static
fi
mkdir -p $BUILD_DIR && cd $BUILD_DIR
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
@@ -21,7 +25,7 @@ build_mivisionx() {
if [ -n "$GPU_ARCHS" ]; then
GPU_TARGETS="$GPU_ARCHS"
else
GPU_TARGETS="gfx908;gfx90a;gfx940;gfx941;gfx942;gfx1030;gfx1100"
GPU_TARGETS="gfx908;gfx90a;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201"
fi
cmake \
@@ -40,8 +44,7 @@ build_mivisionx() {
cpack -G ${PKGTYPE^^}
rm -rf _CPack_Packages/ && find -name '*.o' -delete
mkdir -p $PACKAGE_DIR
cp ${BUILD_DIR}/*.${PKGTYPE} $PACKAGE_DIR
mkdir -p $PACKAGE_DIR && cp ${BUILD_DIR}/*.${PKGTYPE} $PACKAGE_DIR
show_build_cache_stats
}

View File

@@ -1,141 +0,0 @@
#!/bin/bash
source "$(dirname "${BASH_SOURCE}")/compute_utils.sh"
PROJ_NAME=OpenCL-ICD-Loader
TARGET="build"
MAKEOPTS="$DASH_JAY"
BUILD_TYPE="Debug"
PACKAGE_ROOT="$(getPackageRoot)"
PACKAGE_DEB="$PACKAGE_ROOT/deb/${PROJ_NAME,,}"
PACKAGE_RPM="$PACKAGE_ROOT/rpm/${PROJ_NAME,,}"
CLEAN_OR_OUT=0;
PKGTYPE="deb"
MAKETARGET="deb"
API_NAME="rocm-opencl-icd-loader"
printUsage() {
echo
echo "Usage: $(basename "${BASH_SOURCE}") [options ...]"
echo
echo "Options:"
echo " -c, --clean Clean output and delete all intermediate work"
echo " -p, --package <type> Specify packaging format"
echo " -r, --release Make a release build instead of a debug build"
echo " -h, --help Prints this help"
echo " -o, --outdir Print path of output directory containing packages"
echo " -s, --static Component/Build does not support static builds just accepting this param & ignore. No effect of the param on this build"
echo
echo "Possible values for <type>:"
echo " deb -> Debian format (default)"
echo " rpm -> RPM format"
echo
return 0
}
RET_CONFLICT=1
check_conflicting_options $CLEAN_OR_OUT $PKGTYPE $MAKETARGET
if [ $RET_CONFLICT -ge 30 ]; then
print_vars $TARGET $BUILD_TYPE $CLEAN_OR_OUT $PKGTYPE $MAKETARGET
exit $RET_CONFLICT
fi
clean_opencl_icd_loader() {
echo "Cleaning $PROJ_NAME"
rm -rf "$PACKAGE_DEB"
rm -rf "$PACKAGE_RPM"
rm -rf "$PACKAGE_ROOT/${PROJ_NAME,,}"
}
copy_pkg_files_to_rocm() {
local comp_folder=$1
local comp_pkg_name=$2
cd "${OUT_DIR}/${PKGTYPE}/${comp_folder}"|| exit 2
if [ "${PKGTYPE}" = 'deb' ]; then
dpkg-deb -x ${comp_pkg_name}_*.deb pkg/
else
mkdir pkg && pushd pkg/ || exit 2
if [[ "${comp_pkg_name}" != *-dev* ]]; then
rpm2cpio ../${comp_pkg_name}-*.rpm | cpio -idmv
else
rpm2cpio ../${comp_pkg_name}el-*.rpm | cpio -idmv
fi
popd || exit 2
fi
ls ./pkg -alt
cp -r ./pkg/*/rocm*/* "${ROCM_PATH}" || exit 2
rm -rf pkg/
}
build_opencl_icd_loader() {
echo "Downloading $PROJ_NAME" package
if [ "$DISTRO_NAME" = ubuntu ]; then
mkdir -p "$PACKAGE_DEB"
local rocm_ver=${ROCM_VERSION}
if [ ${ROCM_VERSION##*.} = 0 ]; then
rocm_ver=${ROCM_VERSION%.*}
fi
local url="https://repo.radeon.com/rocm/apt/${rocm_ver}/pool/main/r/${API_NAME}/"
local package
package=$(curl -s "$url" | grep -Po 'href="\K[^"]*' | grep "${DISTRO_RELEASE}" | head -n 1)
if [ -z "$package" ]; then
echo "No package found for Ubuntu version $DISTRO_RELEASE"
exit 1
fi
wget -t3 -P "$PACKAGE_DEB" "${url}${package}"
copy_pkg_files_to_rocm ${PROJ_NAME,,} ${API_NAME}
else
echo "$DISTRO_ID is not supported..."
exit 2
fi
echo "Installing $PROJ_NAME" package
}
print_output_directory() {
case ${PKGTYPE} in
("deb")
echo ${PACKAGE_DEB};;
("rpm")
echo ${PACKAGE_RPM};;
(*)
echo "Invalid package type \"${PKGTYPE}\" provided for -o" >&2; exit 1;;
esac
exit
}
VALID_STR=`getopt -o hcraswlo:p: --long help,clean,release,outdir:,package: -- "$@"`
eval set -- "$VALID_STR"
while true ;
do
case "$1" in
(-c | --clean )
TARGET="clean" ; ((CLEAN_OR_OUT|=1)) ; shift ;;
(-r | --release )
BUILD_TYPE="RelWithDebInfo" ; shift ;;
(-h | --help )
printUsage ; exit 0 ;;
(-a | --address_sanitizer)
ack_and_ignore_asan ; shift ;;
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 2 ;;
(-p | --package)
MAKETARGET="$2" ; shift 2;;
(-s | --static)
echo "-s parameter accepted but ignored" ; shift ;;
--) shift; break;;
(*)
echo " This should never come but just incase : UNEXPECTED ERROR Parm : [$1] ">&2 ; exit 20;;
esac
done
case $TARGET in
(clean) clean_opencl_icd_loader ;;
(build) build_opencl_icd_loader ;;
(outdir) print_output_directory ;;
(*) die "Invalid target $TARGET" ;;
esac
echo "Operation complete"

View File

@@ -21,14 +21,15 @@ printUsage() {
return 0
}
PROJ_NAME="opencl-on-rocclr"
MAKEOPTS="$DASH_JAY"
BUILD_PATH="$(getBuildPath opencl-on-rocclr)"
BUILD_PATH="$(getBuildPath $PROJ_NAME)"
TARGET="build"
PACKAGE_ROOT="$(getPackageRoot)"
PACKAGE_DEB="$PACKAGE_ROOT/deb/opencl-on-rocclr"
PACKAGE_RPM="$PACKAGE_ROOT/rpm/opencl-on-rocclr"
PACKAGE_DEB="$PACKAGE_ROOT/deb/$PROJ_NAME"
PACKAGE_RPM="$PACKAGE_ROOT/rpm/$PROJ_NAME"
CORE_BUILD_DIR="$(getBuildPath hsa-core)"
ROCclr_BUILD_DIR="$(getBuildPath rocclr)"
BUILD_TYPE="Debug"
@@ -54,7 +55,7 @@ do
set_asan_env_vars
set_address_sanitizer_on ; shift ;;
(-s | --static)
SHARED_LIBS="OFF" ; shift ;;
ack_and_skip_static ;;
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 2 ;;
--) shift; break;;
@@ -148,7 +149,7 @@ print_output_directory() {
case $TARGET in
(clean) clean_opencl_on_rocclr ;;
(build) build_opencl_on_rocclr ; package_opencl_on_rocclr ;;
(outdir) print_output_directory ;;
(outdir) print_output_directory ;;
(*) die "Invalid target $TARGET" ;;
esac

View File

@@ -13,6 +13,7 @@ printUsage() {
echo " -a, --address_sanitizer Enable address sanitizer"
echo " -o, --outdir <pkg_type> Print path of output directory containing packages of
type referred to by pkg_type"
echo " -s, --static Component/Build does not support static builds just accepting this param for configuring package deps"
echo " -h, --help Prints this help"
echo
echo "Possible values for <type>:"
@@ -23,20 +24,25 @@ printUsage() {
return 0
}
packageMajorVersion="17.60"
PROJ_NAME="openmp-extras"
packageMajorVersion="18.63"
packageMinorVersion="0"
packageVersion="${packageMajorVersion}.${packageMinorVersion}.${ROCM_LIBPATCH_VERSION}"
BUILD_PATH="$(getBuildPath openmp-extras)"
DEB_PATH="$(getDebPath openmp-extras)"
RPM_PATH="$(getRpmPath openmp-extras)"
BUILD_PATH="$(getBuildPath $PROJ_NAME)"
DEB_PATH="$(getDebPath $PROJ_NAME)"
RPM_PATH="$(getRpmPath $PROJ_NAME)"
TARGET="build"
MAKEOPTS="$DASH_JAY"
STATIC_PKG_DEPS="OFF"
export INSTALL_PREFIX=${ROCM_INSTALL_PATH}
while [ "$1" != "" ];
VALID_STR=`getopt -o hcraso:p: --long help,clean,release,address_sanitizer,static,outdir,package: -- "$@"`
eval set -- "$VALID_STR"
while true ;
do
case $1 in
case "$1" in
-c | --clean )
TARGET="clean" ;;
-p | --package )
@@ -52,8 +58,11 @@ do
export SANITIZER=1 ;;
-o | --outdir )
shift 1; PKGTYPE=$1 ; TARGET="outdir" ;;
-s | --static )
export STATIC_PKG_DEPS="ON" ;;
-h | --help )
printUsage ; exit 0 ;;
--) shift; break;;
*)
MAKEARG=$@ ; break ;;
esac
@@ -124,6 +133,23 @@ build_openmp_extras() {
export AOMP_JENKINS_BUILD_LIST="extras openmp pgmath flang flang_runtime"
echo "BEGIN Build of openmp-extras"
"$AOMP_REPOS"/aomp/bin/build_aomp.sh $MAKEARG
local llvm_ver=`$INSTALL_PREFIX/lib/llvm/bin/clang --print-resource-dir | sed 's^/llvm/lib/clang/^ ^' | awk '{print $2}'`
if [ ! -e $ROCM_INSTALL_PATH/lib/llvm/lib/clang/$llvm_ver/include/omp.h ] ; then
if [ ! -h $ROCM_INSTALL_PATH/lib/llvm/lib/clang/$llvm_ver/include/omp.h ] ; then
ln -s ../../../../include/omp.h $ROCM_INSTALL_PATH/lib/llvm/lib/clang/$llvm_ver/include/omp.h
fi
fi
if [ ! -e $ROCM_INSTALL_PATH/lib/llvm/lib/clang/$llvm_ver/include/ompt.h ] ; then
if [ ! -h $ROCM_INSTALL_PATH/lib/llvm/lib/clang/$llvm_ver/include/ompt.h ] ; then
ln -s ../../../../include/ompt.h $ROCM_INSTALL_PATH/lib/llvm/lib/clang/$llvm_ver/include/ompt.h
fi
fi
if [ ! -e $ROCM_INSTALL_PATH/lib/llvm/lib/clang/$llvm_ver/include/omp-tools.h ] ; then
if [ ! -h $ROCM_INSTALL_PATH/lib/llvm/lib/clang/$llvm_ver/include/omp-tools.h ] ; then
ln -s ../../../../include/omp-tools.h $ROCM_INSTALL_PATH/lib/llvm/lib/clang/$llvm_ver/include/omp-tools.h
fi
fi
popd
}
@@ -133,20 +159,30 @@ package_openmp_extras_deb() {
local packageArch="amd64"
local packageMaintainer="Openmp Extras Support <openmp-extras.support@amd.com>"
local packageSummary="OpenMP Extras provides openmp and flang libraries."
local packageSummaryLong="openmp-extras $packageVersion is based on LLVM 15 and is used for offloading to Radeon GPUs."
local packageSummaryLong="openmp-extras $packageVersion is based on LLVM 17 and is used for offloading to Radeon GPUs."
local debDependencies="rocm-llvm, rocm-device-libs, rocm-core"
local debRecommends="gcc, g++"
local controlFile="$packageDeb/openmp-extras/DEBIAN/control"
if [ "$packageName" == "openmp-extras-runtime" ]; then
packageType="runtime"
debDependencies="rocm-core, hsa-rocr"
if [ "$STATIC_PKG_DEPS" == "OFF" ]; then
debDependencies="rocm-core, hsa-rocr"
else
echo "static package dependency configuration for runtime" ;
debDependencies="rocm-core, hsa-rocr-static-dev"
fi
else
local debProvides="openmp-extras"
local debConflicts="openmp-extras"
local debReplaces="openmp-extras"
packageType="devel"
debDependencies="$debDependencies, openmp-extras-runtime, hsa-rocr-dev"
if [ "$STATIC_PKG_DEPS" == "OFF" ]; then
debDependencies="$debDependencies, openmp-extras-runtime, hsa-rocr-dev"
else
echo "Enabled static package dependency configuration for dev" ;
debDependencies="$debDependencies, openmp-extras-runtime, hsa-rocr-static-dev"
fi
fi
if [ -f "$BUILD_PATH"/build/installed_files.txt ] && [ ! -d "$INSTALL_PREFIX"/openmp-extras/devel ]; then
@@ -209,6 +245,9 @@ package_openmp_extras_deb() {
cp -r "$AOMP_REPOS"/aomp/examples/fortran "$packageDeb"/openmp-extras"$copyPath"/share/openmp-extras/examples
cp -r "$AOMP_REPOS"/aomp/examples/openmp "$packageDeb"/openmp-extras"$copyPath"/share/openmp-extras/examples
cp -r "$AOMP_REPOS"/aomp/examples/tools "$packageDeb"/openmp-extras"$copyPath"/share/openmp-extras/examples
if [ -e "$AOMP_REPOS/aomp/examples/Makefile.help" ]; then
cp "$AOMP_REPOS"/aomp/examples/Makefile* "$packageDeb"/openmp-extras"$copyPath"/share/openmp-extras/examples
fi
clean_examples "$packageDeb"/openmp-extras"$copyPath"/share/openmp-extras/examples
fi
@@ -260,7 +299,7 @@ package_openmp_extras_asan_deb() {
local packageArch="amd64"
local packageMaintainer="Openmp Extras Support <openmp-extras.support@amd.com>"
local packageSummary="AddressSanitizer OpenMP Extras provides instrumented openmp and flang libraries."
local packageSummaryLong="openmp-extras $packageVersion is based on LLVM 15 and is used for offloading to Radeon GPUs."
local packageSummaryLong="openmp-extras $packageVersion is based on LLVM 17 and is used for offloading to Radeon GPUs."
local debDependencies="hsa-rocr-asan, rocm-core-asan"
local debRecommends="gcc, g++"
local controlFile="$packageDeb/openmp-extras/DEBIAN/control"
@@ -317,23 +356,26 @@ package_openmp_extras_rpm() {
local packageRpm="$packageDir/rpm"
local specFile="$packageDir/$packageName.spec"
local packageSummary="OpenMP Extras provides openmp and flang libraries."
local packageSummaryLong="openmp-extras $packageVersion is based on LLVM 15 and is used for offloading to Radeon GPUs."
local packageSummaryLong="openmp-extras $packageVersion is based on LLVM 17 and is used for offloading to Radeon GPUs."
local rpmRequires="rocm-llvm, rocm-device-libs, rocm-core"
if [ "$packageName" == "openmp-extras-runtime" ]; then
packageType="runtime"
rpmRequires="rocm-core, hsa-rocr"
if [ "$STATIC_PKG_DEPS" == "OFF" ]; then
rpmRequires="rocm-core, hsa-rocr"
else
rpmRequires="rocm-core, hsa-rocr-static-devel"
fi
else
local rpmProvides="openmp-extras"
local rpmObsoletes="openmp-extras"
packageType="devel"
rpmRequires="$rpmRequires, openmp-extras-runtime, hsa-rocr-devel"
if [ "$STATIC_PKG_DEPS" == "OFF" ]; then
rpmRequires="$rpmRequires, openmp-extras-runtime, hsa-rocr-devel"
else
rpmRequires="$rpmRequires, openmp-extras-runtime, hsa-rocr-static-devel"
fi
fi
rm -f "$AOMP_REPOS"/aomp/examples/*.sh
rm -f "$AOMP_REPOS"/aomp/examples/fortran/*.sh
rm -f "$AOMP_REPOS"/aomp/examples/openmp/*.sh
if [ "$packageType" == "runtime" ]; then
rm -rf "$packageDir"
rm -rf "$RPM_PATH"
@@ -354,6 +396,7 @@ package_openmp_extras_rpm() {
echo "Group: System Environment/Libraries"
echo "License: MIT and ASL 2.0 and ASL 2.0 with exceptions"
echo "Vendor: Advanced Micro Devices, Inc."
echo "Prefix: $INSTALL_PREFIX"
echo "Requires: $rpmRequires"
echo "%if %is_devel"
echo "Provides: $rpmProvides"
@@ -435,6 +478,9 @@ package_openmp_extras_rpm() {
echo " cp -r $AOMP_REPOS/aomp/examples/fortran \$RPM_BUILD_ROOT$copyPath/share/openmp-extras/examples"
echo " cp -r $AOMP_REPOS/aomp/examples/openmp \$RPM_BUILD_ROOT$copyPath/share/openmp-extras/examples"
echo " cp -r $AOMP_REPOS/aomp/examples/tools \$RPM_BUILD_ROOT$copyPath/share/openmp-extras/examples"
if [ -e "$AOMP_REPOS/aomp/examples/Makefile.help" ]; then
echo " cp $AOMP_REPOS/aomp/examples/Makefile* \$RPM_BUILD_ROOT$copyPath/share/openmp-extras/examples"
fi
clean_examples \$RPM_BUILD_ROOT$copyPath/share/openmp-extras/examples
echo "%endif"
echo "%clean"
@@ -461,7 +507,7 @@ package_openmp_extras_asan_rpm() {
local packageRpm="$packageDir/rpm"
local specFile="$packageDir/$packageName.spec"
local packageSummary="AddressSanitizer OpenMP Extras provides instrumented openmp and flang libraries."
local packageSummaryLong="openmp-extras $packageVersion is based on LLVM 15 and is used for offloading to Radeon GPUs."
local packageSummaryLong="openmp-extras $packageVersion is based on LLVM 17 and is used for offloading to Radeon GPUs."
local rpmRequires="hsa-rocr-asan, rocm-core-asan"
local asanLibDir="runtime"
@@ -527,7 +573,6 @@ package_openmp_extras_asan_rpm() {
mv $packageRpm/RPMS/x86_64/*.rpm $RPM_PATH
}
package_openmp_extras() {
local DISTRO_NAME=$(cat /etc/os-release | grep -e ^NAME=)
local installPath="$ROCM_INSTALL_PATH/lib/llvm"
@@ -563,21 +608,23 @@ package_tests_deb(){
local packageArch="amd64"
local packageMaintainer="Openmp Extras Support <openmp-extras.support@amd.com>"
local packageSummary="Tests for openmp-extras."
local packageSummaryLong="Tests for openmp-extras $packageMajorVersion-$packageMinorVersion is based on LLVM 15 and is used for offloading to Radeon GPUs."
local debDependencies="rocm-core"
local debRecommends="gcc, g++"
local packageSummaryLong="Tests for openmp-extras $packageMajorVersion-$packageMinorVersion is based on LLVM 17 and is used for offloading to Radeon GPUs."
local debDependencies="openmp-extras-dev, rocm-core"
local debRecommends=""
local controlFile="$packageDeb/openmp-extras/DEBIAN/control"
local installPath="$ROCM_INSTALL_PATH/share/openmp-extras/tests"
local packageName="openmp-extras-tests"
rm -rf "$packageDir"
mkdir -p $packageDeb/openmp-extras$installPath; mkdir -p $packageDeb/openmp-extras$copyPath/bin
mkdir -p $packageDeb/openmp-extras"$installPath"
if [ -e $(dirname $controlFile) ]; then
rm $(dirname $controlFile)
fi
mkdir -p "$(dirname $controlFile)"
cp -r "$AOMP_REPOS/aomp/test/smoke" "$packageDeb$installPath"
cp -r "$AOMP_REPOS/aomp/." "$packageDeb/openmp-extras/$installPath"
rm -rf "$packageDeb"/openmp-extras"$installPath"/.git "$packageDeb"/openmp-extras"$installPath"/.github
cp "$OUT_DIR/build/lightning/bin/FileCheck" "$packageDeb/openmp-extras/$installPath/bin"
{
echo "Package: $packageName"
echo "Architecture: $packageArch"
@@ -603,11 +650,12 @@ package_tests_rpm(){
local packageName="openmp-extras-tests"
local specFile="$packageDir/$packageName.spec"
local packageSummary="Tests for openmp-extras."
local packageSummaryLong="Tests for openmp-extras $packageVersion is based on LLVM 15 and is used for offloading to Radeon GPUs."
local packageSummaryLong="Tests for openmp-extras $packageVersion is based on LLVM 18 and is used for offloading to Radeon GPUs."
rm -rf "$packageDir"
mkdir -p "$packageRpm$installPath"
mkdir -p "$packageRpm/openmp-extras/$installPath"
{
echo "AutoReqProv: no"
echo "Name: $packageName"
echo "Version: $packageVersion"
echo "Release: ${CPACK_RPM_PACKAGE_RELEASE}%{?dist}"
@@ -615,7 +663,10 @@ package_tests_rpm(){
echo "Group: System Environment/Libraries"
echo "License: Advanced Micro Devices, Inc."
echo "Vendor: Advanced Micro Devices, Inc."
echo "Prefix: $INSTALL_PREFIX"
echo "Requires: $rpmRequires"
echo "%define debug_package %{nil}"
# Redefining __os_install_post to remove stripping
echo "%define __os_install_post %{nil}"
echo "%description"
echo "$packageSummaryLong"
@@ -625,18 +676,21 @@ package_tests_rpm(){
echo "%build"
echo "%install"
echo "mkdir -p \$RPM_BUILD_ROOT$copyPath/share/aomp/tests"
echo "cp -R $AOMP_REPOS/aomp/test/smoke \$RPM_BUILD_ROOT$copyPath/share/aomp/tests"
echo "mkdir -p \$RPM_BUILD_ROOT$installPath"
echo "cp -R $AOMP_REPOS/aomp/. \$RPM_BUILD_ROOT$installPath"
echo "rm -rf \$RPM_BUILD_ROOT$installPath/.git \$RPM_BUILD_ROOT$installPath/.github"
echo "cp $OUT_DIR/build/lightning/bin/FileCheck \$RPM_BUILD_ROOT$installPath/bin"
echo 'find $RPM_BUILD_ROOT \! -type d | sed "s|$RPM_BUILD_ROOT||"> files.list'
echo "%clean"
echo "rm -rf \$RPM_BUILD_ROOT"
echo "%files -f files.list"
echo "$installPath"
echo "%defattr(-,root,root,-)"
echo "%postun"
echo "rm -rf $ROCM_INSTALL_PATH/share/aomp"
echo "rm -rf $installPath"
} > $specFile
rpmbuild --define "_topdir $packageRpm" -ba $specFile
mv $packageRpm/RPMS/x86_64/*.rpm $RPM_PATH
@@ -665,7 +719,7 @@ print_output_directory() {
case $TARGET in
(clean) clean_openmp_extras ;;
(build) build_openmp_extras; package_openmp_extras ;;
(build) build_openmp_extras; package_openmp_extras; package_tests ;;
(outdir) print_output_directory ;;
(*) die "Invalid target $TARGET" ;;
esac

View File

@@ -10,6 +10,10 @@ ENABLE_ADDRESS_SANITIZER=false
build_rccl() {
echo "Start build"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
ack_and_skip_static
fi
mkdir -p $ROCM_PATH/.info/
echo $ROCM_VERSION | tee $ROCM_PATH/.info/version
@@ -23,7 +27,7 @@ build_rccl() {
if [ -n "$GPU_ARCHS" ]; then
GPU_TARGETS="$GPU_ARCHS"
else
GPU_TARGETS="gfx908:xnack-;gfx90a:xnack-;gfx90a:xnack+;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101"
GPU_TARGETS="gfx908:xnack-;gfx90a:xnack-;gfx90a:xnack+;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201"
fi
init_rocm_common_cmake_params

View File

@@ -13,7 +13,7 @@ printUsage() {
echo " -a, --address_sanitizer Enable address sanitizer"
echo " -o, --outdir <pkg_type> Print path of output directory containing packages of
type referred to by pkg_type"
echo " -s, --static Build static lib (.a). build instead of dynamic/shared(.so) "
echo " -s, --static Component/Build does not support static builds just accepting this param & ignore. No effect of the param on this build"
echo " -h, --help Prints this help"
echo
return 0
@@ -41,7 +41,7 @@ RDC_PKG_NAME_ROOT="rdc"
RDC_PKG_NAME="${RDC_PKG_NAME_ROOT}"
GRPC_PROTOC_ROOT="${RDC_BUILD_DIR}/grpc"
GRPC_SEARCH_ROOT="/usr/grpc"
GRPC_DESIRED_VERSION="1.59.1" # do not include 'v'
GRPC_DESIRED_VERSION="1.61.0"
RDC_LIB_RPATH='$ORIGIN'
RDC_LIB_RPATH=$RDC_LIB_RPATH:'$ORIGIN/..'
@@ -70,7 +70,7 @@ do
(-d | --documentation )
BUILD_DOCS="yes" ;;
(-s | --static)
SHARED_LIBS="OFF" ; shift ;;
ack_and_skip_static ;;
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 2 ;;
(-p | --package)
@@ -111,49 +111,20 @@ find_grpc() {
GRPC_PROTOC_ROOT=$GRPC_SEARCH_ROOT
}
build_grpc() {
if find_grpc; then
return 0
fi
echo "GRPC SEARCH FAILED! Building from scratch..."
mkdir -p $PACKAGE_ROOT/build
pushd $PACKAGE_ROOT/build
if [ ! -d $PACKAGE_ROOT/build/grpc/.git ]; then
git clone \
--shallow-submodules \
--recurse-submodules \
$DASH_JAY \
-b v${GRPC_DESIRED_VERSION} \
--depth 1 \
https://github.com/grpc/grpc
fi
cd grpc
mkdir -p cmake/build
cd cmake/build
cmake \
-DgRPC_INSTALL=ON \
-DgRPC_BUILD_TESTS=OFF \
-DBUILD_SHARED_LIBS=ON \
-DCMAKE_BUILD_TYPE=Release \
-DCMAKE_INSTALL_PREFIX=${GRPC_PROTOC_ROOT} \
../..
cmake --build . -- $DASH_JAY
cmake --build . -- install
cp ../../LICENSE ${GRPC_PROTOC_ROOT}
popd
}
rdc_backwards_compat_cmake_params() {
grep -q "RDC_CLIENT_INSTALL_PREFIX" "$RDC_ROOT/CMakeLists.txt" &&
echo "-DRDC_CLIENT_INSTALL_PREFIX=$PACKAGE_ROOT"
}
build_rdc() {
if ! find_grpc; then
echo "ERROR: GRPC SEARCH FAILED!"
echo "You are expected to have gRPC [${GRPC_DESIRED_VERSION}] in [${GRPC_SEARCH_ROOT}]"
# Compiling gRPC as part of the RDC build takes too long and times out the build job
return 1
fi
echo "gRPC [${GRPC_DESIRED_VERSION}] found!"
echo "Building RDC"
echo "RDC_BUILD_DIR: ${RDC_BUILD_DIR}"
echo "GRPC_PROTOC_ROOT: ${GRPC_PROTOC_ROOT}"
@@ -226,7 +197,7 @@ verifyEnvSetup
case $TARGET in
(clean) clean_rdc ;;
(clean_grpc) clean_grpc ;;
(build) build_grpc; build_rdc ;;
(build) build_rdc ;;
(outdir) print_output_directory ;;
(*) die "Invalid target $TARGET" ;;
esac

View File

@@ -7,11 +7,6 @@ set_component_src rocAL
build_rocal() {
if [ "$DISTRO_ID" = "mariner-2.0" ] ; then
echo "Not building rocal for ${DISTRO_ID}. Exiting..."
return 0
fi
echo "Start build"
# Enable ASAN

View File

@@ -10,6 +10,10 @@ set_component_src rocALUTION
build_rocalution() {
echo "Start build"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
ack_and_skip_static
fi
cd $COMPONENT_SRC
CXX="g++"
@@ -27,7 +31,7 @@ build_rocalution() {
if [ -n "$GPU_ARCHS" ]; then
GPU_TARGETS="$GPU_ARCHS"
else
GPU_TARGETS="gfx908:xnack-;gfx90a:xnack-;gfx90a:xnack+;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101"
GPU_TARGETS="gfx908:xnack-;gfx90a:xnack-;gfx90a:xnack+;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201"
fi
cmake \
@@ -35,7 +39,6 @@ build_rocalution() {
${LAUNCHER_FLAGS} \
"${rocm_math_common_cmake_params[@]}" \
-DAMDGPU_TARGETS=${GPU_TARGETS} \
-DCPACK_SET_DESTDIR=OFF \
-DBUILD_CLIENTS_SAMPLES=ON \
-DBUILD_CLIENTS_TESTS=ON \
-DBUILD_CLIENTS_BENCHMARKS=ON \

View File

@@ -12,6 +12,11 @@ stage2_command_args "$@"
build_rocblas() {
echo "Start build"
SHARED_LIBS="ON"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
SHARED_LIBS="OFF"
fi
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
set_address_sanitizer_on
@@ -26,7 +31,7 @@ build_rocblas() {
if [ -n "$GPU_ARCHS" ]; then
GPU_TARGETS="$GPU_ARCHS"
else
GPU_TARGETS="gfx908:xnack-;gfx90a:xnack+;gfx90a:xnack-;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101"
GPU_TARGETS="gfx900;gfx906:xnack-;gfx908:xnack-;gfx90a:xnack+;gfx90a:xnack-;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201"
fi
init_rocm_common_cmake_params
@@ -36,8 +41,8 @@ build_rocblas() {
"${rocm_math_common_cmake_params[@]}" \
-DROCM_DIR="${ROCM_PATH}" \
${LAUNCHER_FLAGS} \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
-DCMAKE_PREFIX_PATH="${DEPS_DIR};${ROCM_PATH}" \
-DCPACK_SET_DESTDIR=OFF \
-DBUILD_CLIENTS_TESTS=ON \
-DBUILD_CLIENTS_BENCHMARKS=ON \
-DBUILD_CLIENTS_SAMPLES=ON \

View File

@@ -1,127 +0,0 @@
#!/bin/bash
source "$(dirname "${BASH_SOURCE}")/compute_utils.sh"
printUsage() {
echo
echo "Usage: $(basename "${BASH_SOURCE}") [options ...] [make options]"
echo
echo "Options:"
echo " -h, --help Prints this help"
echo " -c, --clean Clean output and delete all intermediate work"
echo " -r, --release Make a release build instead of a debug build"
echo " -a, --address_sanitizer Enable address sanitizer"
echo " -s, --static Build static lib (.a). build instead of dynamic/shared(.so) "
echo " -o, --outdir <pkg_type> Print path of output directory containing packages of type referred to by pkg_type"
echo
echo "Possible values for <type>:"
echo " deb -> Debian format (default)"
echo " rpm -> RPM format"
echo
return 0
}
MAKEOPTS="$DASH_JAY"
BUILD_PATH="$(getBuildPath rocclr)"
TARGET="build"
PACKAGE_ROOT="$(getPackageRoot)"
PACKAGE_DEB="$(getPackageRoot)/deb/rocclr"
PACKAGE_RPM="$(getPackageRoot)/rpm/rocclr"
CORE_BUILD_DIR="$(getBuildPath hsa-core)"
BUILD_TYPE="Debug"
SHARED_LIBS="ON"
CLEAN_OR_OUT=0;
MAKETARGET="deb"
PKGTYPE="deb"
VALID_STR=`getopt -o hcraso: --long help,clean,release,static,address_sanitizer,outdir: -- "$@"`
eval set -- "$VALID_STR"
while true ;
do
case "$1" in
(-h | --help)
printUsage ; exit 0;;
(-c | --clean)
TARGET="clean" ; ((CLEAN_OR_OUT|=1)) ; shift ;;
(-r | --release)
BUILD_TYPE="Release" ; shift ;;
(-a | --address_sanitizer)
set_asan_env_vars
set_address_sanitizer_on ; shift ;;
(-s | --static)
SHARED_LIBS="OFF" ; shift ;;
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 2 ;;
--) shift; break;;
(*)
echo " This should never come but just incase : UNEXPECTED ERROR Parm : [$1] ">&2 ; exit 20;;
esac
done
RET_CONFLICT=1
check_conflicting_options $CLEAN_OR_OUT $PKGTYPE $MAKETARGET
if [ $RET_CONFLICT -ge 30 ]; then
print_vars $API_NAME $TARGET $BUILD_TYPE $SHARED_LIBS $CLEAN_OR_OUT $PKGTYPE $MAKETARGET
exit $RET_CONFLICT
fi
clean_rocclr() {
rm -rf "$BUILD_PATH"
rm -rf "$PACKAGE_DEB"
rm -rf "$PACKAGE_RPM"
}
build_rocclr() {
if [ "$SHARED_LIBS" = "ON" ]; then
echo "rocclr not a standalone repo. skipping build" >&2
echo "rocclr not a standalone repo. skipping build"
exit 0
fi
if [ ! -e "$CLR_ROOT/CMakeLists.txt" ]; then
_ROCclr_CMAKELIST_DIR="$CLR_ROOT"
elif [ ! -e "$ROCclr_ROOT/CMakeLists.txt" ]; then
echo "No $ROCclr_ROOT/CMakeLists.txt file, skipping rocclr" >&2
echo "No $ROCclr_ROOT/CMakeLists.txt file, skipping rocclr"
exit 0
else
_ROCclr_CMAKELIST_DIR="$ROCclr_ROOT"
fi
echo "$_ROCclr_CMAKELIST_DIR"
mkdir -p "$BUILD_PATH"
pushd "$BUILD_PATH"
print_lib_type $SHARED_LIBS
if [ ! -e Makefile ]; then
echo "Building ROCclr CMake environment"
cmake -DUSE_COMGR_LIBRARY=ON \
$(rocm_cmake_params) \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
-DLLVM_INCLUDES="$LLVM_ROOT/include" \
$(rocm_common_cmake_params) \
"$_ROCclr_CMAKELIST_DIR"
echo "CMake complete"
fi
echo "Building ROCclr"
cmake --build . -- $MAKEOPTS "VERBOSE=1"
popd
}
case $TARGET in
(clean) clean_rocclr ;;
(build) build_rocclr ;;
(outdir) exit ;;
(*) die "Invalid target $TARGET" ;;
esac
echo "Operation complete"

View File

@@ -4,14 +4,14 @@ source "$(dirname "${BASH_SOURCE[0]}")/compute_helper.sh"
set_component_src rocDecode
BUILD_DEV=ON
build_rocdecode() {
if [ "$DISTRO_ID" = "centos-7" ] || [ "$DISTRO_ID" = "sles-15.4" ] ; then
echo "Not building rocDecode for ${DISTRO_ID}. Exiting..."
return 0
echo "Start build"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
ack_and_skip_static
fi
echo "Start build"
mkdir -p $BUILD_DIR && cd $BUILD_DIR
python3 ${COMPONENT_SRC}/rocDecode-setup.py --developer OFF
# python3 ${COMPONENT_SRC}/rocDecode-setup.py --developer OFF
cmake -DROCM_DEP_ROCMCORE=ON ${COMPONENT_SRC}
make -j8

View File

@@ -21,7 +21,7 @@ build_rocfft() {
if [ -n "$GPU_ARCHS" ]; then
GPU_TARGETS="$GPU_ARCHS"
else
GPU_TARGETS="gfx908;gfx90a;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101"
GPU_TARGETS="gfx908;gfx90a;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201"
fi
CXX="${ROCM_PATH}/bin/hipcc" \
@@ -34,7 +34,6 @@ build_rocfft() {
-DBUILD_CLIENTS_SAMPLES=ON \
-DBUILD_CLIENTS_TESTS=ON \
-DBUILD_CLIENTS_RIDER=ON \
-DCPACK_SET_DESTDIR=OFF \
"$COMPONENT_SRC"
cmake --build "$BUILD_DIR" -- -j${PROC}

View File

@@ -0,0 +1,39 @@
#!/bin/bash
set -ex
source "$(dirname "${BASH_SOURCE[0]}")/compute_helper.sh"
set_component_src rocJPEG
BUILD_DEV=ON
build_rocjpeg() {
echo "Start build"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
ack_and_skip_static
fi
mkdir -p $BUILD_DIR && cd $BUILD_DIR
# python3 ../rocJPEG-setup.py
cmake -DROCM_DEP_ROCMCORE=ON "$COMPONENT_SRC"
make -j8
make install
make package
cmake --build "$BUILD_DIR" -- -j${PROC}
cpack -G ${PKGTYPE^^}
rm -rf _CPack_Packages/ && find -name '*.o' -delete
mkdir -p $PACKAGE_DIR
cp ${BUILD_DIR}/*.${PKGTYPE} $PACKAGE_DIR
show_build_cache_stats
}
clean_rocjpeg() {
echo "Cleaning rocJPEG build directory: ${BUILD_DIR} ${PACKAGE_DIR}"
rm -rf "$BUILD_DIR" "$PACKAGE_DIR"
echo "Done!"
}
stage2_command_args "$@"
case $TARGET in
build) build_rocjpeg ;;
outdir) print_output_directory ;;
clean) clean_rocjpeg ;;
*) die "Invalid target $TARGET" ;;
esac

View File

@@ -32,7 +32,6 @@ ROCM_CMAKE_BUILD_DIR="$(getBuildPath rocm-cmake)"
ROCM_CMAKE_BUILD_DIR="$(getBuildPath rocm-cmake)"
ROCM_CMAKE_PACKAGE_DEB="$(getPackageRoot)/deb/rocm-cmake"
ROCM_CMAKE_PACKAGE_RPM="$(getPackageRoot)/rpm/rocm-cmake"
ROCM_WHEEL_DIR="${ROCM_CMAKE_BUILD_DIR}/_wheel"
ROCM_CMAKE_BUILD_TYPE="debug"
BUILD_TYPE="Debug"
SHARED_LIBS="ON"
@@ -56,8 +55,6 @@ do
ack_and_ignore_asan ; shift ;;
(-s | --static)
SHARED_LIBS="OFF" ; shift ;;
(-w | --wheel)
WHEEL_PACKAGE=true ; shift ;;
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 2 ;;
(-p | --package)
@@ -78,7 +75,6 @@ fi
clean_rocm_cmake() {
rm -rf "$ROCM_WHEEL_DIR"
rm -rf $ROCM_CMAKE_BUILD_DIR
rm -rf $ROCM_CMAKE_PACKAGE_DEB
rm -rf $ROCM_CMAKE_PACKAGE_RPM
@@ -106,19 +102,6 @@ build_rocm_cmake() {
copy_if RPM "${CPACKGEN:-"DEB;RPM"}" "$ROCM_CMAKE_PACKAGE_RPM" $ROCM_CMAKE_BUILD_DIR/rocm-cmake*.rpm
}
create_wheel_package() {
echo "Creating rocm-cmake wheel package"
# Copy the setup.py generator to build folder
mkdir -p $ROCM_WHEEL_DIR
cp -f $SCRIPT_ROOT/generate_setup_py.py $ROCM_WHEEL_DIR
cp -f $SCRIPT_ROOT/repackage_wheel.sh $ROCM_WHEEL_DIR
cd $ROCM_WHEEL_DIR
# Currently only supports python3.6
./repackage_wheel.sh $ROCM_CMAKE_BUILD_DIR/rocm-cmake*.rpm python3.6
# Copy the wheel created to RPM folder which will be uploaded to artifactory
copy_if WHL "WHL" "$ROCM_CMAKE_PACKAGE_RPM" "$ROCM_WHEEL_DIR"/dist/*.whl
}
print_output_directory() {
case ${PKGTYPE} in
("deb")
@@ -138,9 +121,4 @@ case $TARGET in
(*) die "Invalid target $TARGET" ;;
esac
if [[ $WHEEL_PACKAGE == true ]]; then
echo "Wheel Package build started !!!!"
create_wheel_package
fi
echo "Operation complete"

View File

@@ -24,10 +24,11 @@ printUsage() {
return 0
}
PROJ_NAME="rocm-core"
PACKAGE_ROOT="$(getPackageRoot)"
ROCM_CORE_BUILD_DIR="$(getBuildPath rocm_core)"
ROCM_CORE_PACKAGE_DEB="$(getPackageRoot)/deb/rocm-core"
ROCM_CORE_PACKAGE_RPM="$(getPackageRoot)/rpm/rocm-core"
ROCM_CORE_PACKAGE_DEB="$(getPackageRoot)/deb/$PROJ_NAME"
ROCM_CORE_PACKAGE_RPM="$(getPackageRoot)/rpm/$PROJ_NAME"
ROCM_CORE_MAKE_OPTS="$DASH_JAY -C $ROCM_CORE_BUILD_DIR"
BUILD_TYPE="Debug"
TARGET="build"

View File

@@ -48,7 +48,7 @@ CLEAN_OR_OUT=0;
MAKETARGET="deb"
PKGTYPE="deb"
LDFLAGS="$LDFLAGS -Wl,--enable-new-dtags"
LIB_AMD_PYTHON="libamdpython.so"
tokeep=(
main${ROCM_INSTALL_PATH}/bin/rocgdb
@@ -123,11 +123,41 @@ package_deb(){
local VERSION
get_version unknown
VERSION="${VERSION}.${ROCM_LIBPATCH_VERSION}"
grep -v '^# ' > "$BUILD_DIR/package/main/DEBIAN/control" <<EOF
#create postinstall and prerm
grep -v '^# ' > "$BUILD_DIR/package/main/DEBIAN/preinst" <<EOF
#!/bin/sh
# Pre-installation script commands
echo "Running pre-installation script..."
mkdir -p ${ROCM_INSTALL_PATH}/lib
PYTHON_LIB_INSTALLED=\$(ldconfig -p | awk '/libpython3/ { print \$NF; exit}')
ln -s \$PYTHON_LIB_INSTALLED ${ROCM_INSTALL_PATH}/lib/$LIB_AMD_PYTHON
echo "pre-installation done."
EOF
grep -v '^# ' > "$BUILD_DIR/package/main/DEBIAN/postrm" <<EOF
#!/bin/sh
# Post-uninstallation script commands
echo "Running post-uninstallation script..."
PYTHON_LINK_BY_OPENCL=\$(ldconfig -p | awk '/libpython3/ { print \$NF; exit}' | awk -F'/' '{print \$NF}')
rm -f ${ROCM_INSTALL_PATH}/lib/\$PYTHON_LINK_BY_OPENCL
rm -f ${ROCM_INSTALL_PATH}/lib/$LIB_AMD_PYTHON
if [ -L "${ROCM_INSTALL_PATH}/lib/$LIB_AMD_PYTHON" ] || \
[ -L "${ROCM_INSTALL_PATH}/lib/\$PYTHON_LINK_BY_OPENCL" ] ; then
echo " some rocm-gdb requisite libs could not be removed"
else
echo " all requisite libs removed successfully "
fi
echo "post-uninstallation done."
EOF
chmod +x $BUILD_DIR/package/main/DEBIAN/postrm
chmod +x $BUILD_DIR/package/main/DEBIAN/preinst
# Create control file, with variable substitution.
# Lines with # at the start are removed, to allow for comments
mkdir "$BUILD_DIR/debian"
grep -v '^# ' > "$BUILD_DIR/debian/control" <<EOF
# Required fields
Version: ${VERSION}-${CPACK_DEBIAN_PACKAGE_RELEASE}
Package: ${PROJ_NAME}
Source: ${PROJ_NAME}-src
Maintainer: ROCm Debugger Support <rocm-gdb.support@amd.com>
Description: ROCgdb
This is ROCgdb, the AMD ROCm source-level debugger for Linux,
@@ -137,15 +167,37 @@ Section: utils
Architecture: amd64
Essential: no
Priority: optional
Depends: libexpat1, libtinfo5, libncurses5, rocm-dbgapi, libpython3.10 | libpython3.8, libbabeltrace-ctf1 (>= 1.2.1), libbabeltrace1 (>= 1.2.1), rocm-core
Depends: \${shlibs:Depends}, rocm-dbgapi, rocm-core
EOF
# Use dpkg-shlibdeps to list shlib dependencies, the result is placed
# in $BUILD_DIR/debian/substvars.
(
cd "$BUILD_DIR"
if [[ $ASAN_BUILD == "yes" ]]
then
LD_LIBRARY_PATH=${ROCM_INSTALL_PATH}/lib/asan:$LD_LIBRARY_PATH
fi
dpkg-shlibdeps --ignore-missing-info -e "$BUILD_DIR/package/main/${ROCM_INSTALL_PATH}/bin/rocgdb"
)
# Generate the final DEBIAN/control, and substitute the shlibs:Depends.
# This is a bit unorthodox as we are only using bits and pieces of the
# dpkg tools.
(
SHLIB_DEPS=$(grep "^shlibs:Depends" "$BUILD_DIR/debian/substvars" | \
sed -e "s/shlibs:Depends=//")
sed -E \
-e "/^#/d" \
-e "/^Source:/d" \
-e "s/\\$\{shlibs:Depends\}/$SHLIB_DEPS/" \
< debian/control > "$BUILD_DIR/package/main/DEBIAN/control"
)
mkdir -p "$OUT_DIR/deb/$PROJ_NAME"
fakeroot dpkg-deb -Zgzip --build "$BUILD_DIR/package/main" "$OUT_DIR/deb/$PROJ_NAME"
# Package the tests so they can be run on a test slave
mkdir -p "$BUILD_DIR/package/tests/DEBIAN"
mkdir -p "$BUILD_DIR/package/tests/${ROCM_INSTALL_PATH}/test/gdb"
# Create control file, with variable substitution.
# Lines with # at the start are removed, to allow for comments
grep -v '^# ' > "$BUILD_DIR/package/tests/DEBIAN/control" <<EOF
# Required fields
Version: ${VERSION}-${CPACK_DEBIAN_PACKAGE_RELEASE}
@@ -161,7 +213,6 @@ Priority: optional
# rocm-core as policy says everything to depend on rocm-core
Depends: ${PROJ_NAME} (=${VERSION}-${CPACK_DEBIAN_PACKAGE_RELEASE}), dejagnu, rocm-core, make
EOF
copy_testsuite_files
fakeroot dpkg-deb -Zgzip --build "$BUILD_DIR/package/tests" "$OUT_DIR/deb/$PROJ_NAME"
}
@@ -204,7 +255,9 @@ Summary: ROCm source-level debugger for Linux
Version: ${VERSION//-/_}
Release: ${CPACK_RPM_PACKAGE_RELEASE}%{?dist}
License: GPL
Prefix: ${ROCM_INSTALL_PATH}
Requires: rocm-core
Provides: $LIB_AMD_PYTHON()(64bit)
%description
This is ROCgdb, the ROCm source-level debugger for Linux, based on
@@ -225,6 +278,27 @@ https://github.com/RadeonOpenCompute/ROCm
## into the local RPM_BUILD_ROOT and left the defaults take over. Need
## to quote the dollar signs as we want rpm to expand them when it is
## run, rather than the shell when we build the spec file.
%pre
# Post-install script commands
echo "Running post-install script..."
mkdir -p ${ROCM_INSTALL_PATH}/lib
PYTHON_LIB_INSTALLED=\$(ldconfig -p | awk '/libpython3/ { print \$NF; exit}')
ln -s \$PYTHON_LIB_INSTALLED ${ROCM_INSTALL_PATH}/lib/$LIB_AMD_PYTHON
%postun
# Post-uninstallation script commands
echo "Running post-uninstallation script..."
PYTHON_LINK_BY_OPENCL=\$(ldconfig -p | awk '/libpython3/ { print \$NF; exit}' | awk -F'/' '{print \$NF}')
rm -f ${ROCM_INSTALL_PATH}/lib/\$PYTHON_LINK_BY_OPENCL
rm -f ${ROCM_INSTALL_PATH}/lib/$LIB_AMD_PYTHON
if [ -L "${ROCM_INSTALL_PATH}/lib/$LIB_AMD_PYTHON" ] || \
[ -L "${ROCM_INSTALL_PATH}/lib/\$PYTHON_LINK_BY_OPENCL" ] ; then
echo " some rocm-gdb requisite libs could not be removed"
else
echo " all requisite libs removed successfully "
fi
echo "post-uninstallation done."
%install
rm -rf \$RPM_BUILD_ROOT
mkdir -p \$RPM_BUILD_ROOT
@@ -279,6 +353,7 @@ Summary: Tests for gdb enhanced to debug AMD GPUs
Version: ${VERSION//-/_}
Release: ${RELEASE}
License: GPL
Prefix: ${ROCM_INSTALL_PATH}
Requires: dejagnu, ${PROJ_NAME} = ${VERSION//-/_}-${RELEASE}, rocm-core, make
%description
@@ -340,15 +415,36 @@ build() {
--infodir="\${prefix}/share/info/rocgdb" \
--with-separate-debug-dir="\${prefix}/lib/debug:/usr/lib/debug" \
--with-gdb-datadir="\${prefix}/share/rocgdb" --enable-64-bit-bfd \
--with-bugurl="$BUG_URL" --with-pkgversion="${ROCM_BUILD_ID:-ROCm}" \
--enable-targets="x86_64-linux-gnu,amdgcn-amd-amdhsa" \
--disable-ld --disable-gas --disable-gdbserver --disable-sim --enable-tui \
--disable-gdbtk --disable-shared --disable-gprofng \
--with-expat --with-system-zlib --without-guile --with-babeltrace --with-lzma \
--with-python=$pythonver --with-rocm-dbgapi=$ROCM_INSTALL_PATH \
--with-amd-dbgapi PKG_CONFIG_PATH="${ROCM_INSTALL_PATH}/share/pkgconfig" \
--with-bugurl="$BUG_URL" --with-pkgversion="${ROCM_BUILD_ID:-ROCm}" \
--enable-targets="x86_64-linux-gnu,amdgcn-amd-amdhsa" \
--disable-gas \
--disable-gdbserver \
--disable-gdbtk \
--disable-gprofng \
--disable-ld \
--disable-shared \
--disable-sim \
--enable-tui \
--with-amd-dbgapi \
--with-expat \
--with-lzma \
--with-python=$pythonver \
--with-rocm-dbgapi=$ROCM_INSTALL_PATH \
--with-system-zlib \
--with-zstd \
--without-babeltrace \
--without-guile \
--without-intel-pt \
--without-libunwind-ia64 \
--without-xxhash \
PKG_CONFIG_PATH="${ROCM_INSTALL_PATH}/share/pkgconfig" \
LDFLAGS="$LDFLAGS"
LD_RUN_PATH='${ORIGIN}/../lib' make $MAKE_OPTS
REPLACE_LIB_NAME=$(ldd -d $BUILD_DIR/gdb/gdb |awk '/libpython/{print $1}')
echo "Replacing $REPLACE_LIB_NAME with $LIB_AMD_PYTHON"
patchelf --replace-needed $REPLACE_LIB_NAME $LIB_AMD_PYTHON $BUILD_DIR/gdb/gdb
mkdir -p $BUILD_DIR/package/main${ROCM_INSTALL_PATH}/{share/rocgdb,bin}
make $MAKE_OPTS -C gdb DESTDIR=$BUILD_DIR/package/main install install-pdf install-html
@@ -381,7 +477,7 @@ main(){
VALID_STR=`getopt -o hcraso:p: --long help,clean,release,static,address_sanitizer,outdir:,package: -- "$@"`
eval set -- "$VALID_STR"
ASAN_BUILD="no"
while true ;
do
case "$1" in
@@ -393,9 +489,10 @@ do
BUILD_TYPE="Release" ; shift ; MAKEARG="$MAKEARG REL=1" ;; # For compatability with other scripts
(-a | --address_sanitizer)
set_asan_env_vars
set_address_sanitizer_on ; shift ;;
set_address_sanitizer_on
ASAN_BUILD="yes" ; shift ;;
(-s | --static)
SHARED_LIBS="OFF" ; shift ;;
ack_and_skip_static ;;
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 2 ;;
(-p | --package) #FIXME

View File

@@ -7,7 +7,7 @@ printUsage() {
echo "Usage: $(basename "${BASH_SOURCE}") [options ...]"
echo
echo "Options:"
echo " -s, --static Supports static CI by accepting this param & not bailing out. No effect of the param though"
echo " -s, --static Component/Build does not support static builds just accepting this param & ignore. No effect of the param on this build"
echo " -c, --clean Clean output and delete all intermediate work"
echo " -p, --package <type> Specify packaging format"
echo " -r, --release Make a release build instead of a debug build"
@@ -65,7 +65,7 @@ do
set_asan_env_vars
set_address_sanitizer_on ; shift ;;
(-s | --static)
SHARED_LIBS="OFF" ; shift ;;
ack_and_skip_static ;;
(-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 2 ;;
(-p | --package)
@@ -127,6 +127,10 @@ build_rocm_bandwidth_test() {
echo "Packaging $TEST_NAME"
cmake --build "$TEST_BLD_DIR" -- $MAKEARG -C $TEST_BLD_DIR package
mkdir -p "$TEST_BIN_DIR"
echo "Copying $TEST_NAME to $TEST_BIN_DIR"
progressCopy "$TEST_BLD_DIR/$TEST_NAME" "$TEST_BIN_DIR"
copy_if DEB "${CPACKGEN:-"DEB;RPM"}" "$TEST_PKG_DEB" $TEST_BLD_DIR/*.deb
copy_if RPM "${CPACKGEN:-"DEB;RPM"}" "$TEST_PKG_RPM" $TEST_BLD_DIR/*.rpm

View File

@@ -23,6 +23,7 @@ printUsage() {
return 0
}
PROJ_NAME="rsmi"
PACKAGE_ROOT="$(getPackageRoot)"
TARGET="build"
@@ -30,8 +31,8 @@ PACKAGE_LIB=$(getLibPath)
PACKAGE_INCLUDE="$(getIncludePath)"
RSMI_BUILD_DIR=$(getBuildPath rsmi)
RSMI_PACKAGE_DEB_DIR="$(getPackageRoot)/deb/rsmi"
RSMI_PACKAGE_RPM_DIR="$(getPackageRoot)/rpm/rsmi"
RSMI_PACKAGE_DEB_DIR="$(getPackageRoot)/deb/$PROJ_NAME"
RSMI_PACKAGE_RPM_DIR="$(getPackageRoot)/rpm/$PROJ_NAME"
RSMI_BUILD_TYPE="debug"
BUILD_TYPE="Debug"

View File

@@ -22,17 +22,16 @@ printUsage() {
return 0
}
PROJ_NAME="rocminfo"
TARGET="build"
ROCMINFO_DEST="$(getBinPath)"
ROCMINFO_SRC_ROOT="$ROCMINFO_ROOT"
ROCMINFO_BUILD_DIR="$(getBuildPath rocminfo)"
ROCMINFO_BUILD_DIR="$(getBuildPath $PROJ_NAME)"
MAKEARG="$DASH_JAY"
PACKAGE_ROOT="$(getPackageRoot)"
PACKAGE_UTILS="$(getUtilsPath)"
ROCMINFO_PACKAGE_DEB="$(getPackageRoot)/deb/rocminfo"
ROCMINFO_PACKAGE_RPM="$(getPackageRoot)/rpm/rocminfo"
ROCMINFO_PACKAGE_DEB="$PACKAGE_ROOT/deb/$PROJ_NAME"
ROCMINFO_PACKAGE_RPM="$PACKAGE_ROOT/rpm/$PROJ_NAME"
BUILD_TYPE="debug"
SHARED_LIBS="ON"
CLEAN_OR_OUT=0;
@@ -91,6 +90,7 @@ build_rocminfo() {
cmake \
$(rocm_cmake_params) \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
-DROCRTST_BLD_TYPE="$BUILD_TYPE" \
$(rocm_common_cmake_params) \
-DCPACK_PACKAGE_VERSION_MAJOR="1" \

View File

@@ -10,6 +10,10 @@ ROCM_RVS_LIB_RPATH="\$ORIGIN/.."
build_rocmvalidationsuite() {
echo "Start build"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
ack_and_skip_static
fi
if [ "${ENABLE_ADDRESS_SANITIZER}" == "true" ]; then
set_asan_env_vars
set_address_sanitizer_on
@@ -17,11 +21,13 @@ build_rocmvalidationsuite() {
cd "${COMPONENT_SRC}"
mkdir -p "$BUILD_DIR"
init_rocm_common_cmake_params
cmake \
$(rocm_common_cmake_params) \
"${rocm_math_common_cmake_params[@]}" \
-DFETCH_ROCMPATH_FROM_ROCMCORE=ON \
-DCMAKE_SHARED_LINKER_FLAGS_INIT="-Wl,--enable-new-dtags,--rpath,$ROCM_LIB_RPATH:$ROCM_RVS_LIB_RPATH" \
-DCMAKE_SHARED_LINKER_FLAGS_INIT="-Wl,--enable-new-dtags,--build-id=sha1,--rpath,$ROCM_LIB_RPATH:$ROCM_RVS_LIB_RPATH" \
-DRVS_BUILD_TESTS=FALSE \
-B "$BUILD_DIR" \
"$COMPONENT_SRC"

View File

@@ -16,12 +16,17 @@ build_rocprim() {
ASAN_CMAKE_PARAMS="false"
fi
SHARED_LIBS="ON"
if [ "${ENABLE_STATIC_BUILDS}" == "true" ]; then
SHARED_LIBS="OFF"
fi
mkdir -p "$BUILD_DIR" && cd "$BUILD_DIR"
if [ -n "$GPU_ARCHS" ]; then
GPU_TARGETS="$GPU_ARCHS"
else
GPU_TARGETS="gfx908:xnack-;gfx90a:xnack-;gfx90a:xnack+;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101"
GPU_TARGETS="gfx908:xnack-;gfx90a:xnack-;gfx90a:xnack+;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201"
fi
init_rocm_common_cmake_params
@@ -31,8 +36,8 @@ build_rocprim() {
"${rocm_math_common_cmake_params[@]}" \
-DAMDGPU_TARGETS=${GPU_TARGETS} \
-DBUILD_BENCHMARK=OFF \
-DBUILD_SHARED_LIBS=ON \
-DBUILD_TEST=ON \
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
-DCMAKE_MODULE_PATH="${ROCM_PATH}/lib/cmake/hip;${ROCM_PATH}/hip/cmake" \
"$COMPONENT_SRC"

View File

@@ -7,18 +7,18 @@ printUsage() {
echo "Usage: ${BASH_SOURCE##*/} [options ...]"
echo
echo "Options:"
echo " -c, --clean Clean output and delete all intermediate work"
echo " -s, --static Build static lib (.a). build instead of dynamic/shared(.so) "
echo " -p, --package <type> Specify packaging format"
echo " -r, --release Make a release build instead of a debug build"
echo " -h, --help Prints this help"
echo " -a, --address_sanitizer Enable address sanitizer"
echo " -c, --clean Clean output and delete all intermediate work"
echo " -o, --outdir <pkg_type> Print path of output directory containing packages of
type referred to by pkg_type"
echo " -w, --wheel Creates python wheel package of omniperf.
echo " -p, --package <type> Specify packaging format"
echo " -r, --release Make a release build instead of a debug build"
echo " -s, --static Build static lib (.a). build instead of dynamic/shared(.so) "
echo " -w, --wheel Creates python wheel package of ROCm Compute Profiler.
It needs to be used along with -r option"
echo " -h, --help Prints this help"
echo
echo "Possible values for <type>:"
echo "Possible values for package <type>:"
echo " deb -> Debian format (default)"
echo " rpm -> RPM format"
echo
@@ -26,7 +26,7 @@ printUsage() {
return 0
}
API_NAME="omniperf"
API_NAME="rocprofiler-compute"
PROJ_NAME="$API_NAME"
LIB_NAME="lib${API_NAME}"
TARGET="build"
@@ -36,17 +36,13 @@ PACKAGE_LIB="$(getLibPath)"
BUILD_DIR="$(getBuildPath $API_NAME)"
PACKAGE_DEB="$(getPackageRoot)/deb/$API_NAME"
PACKAGE_RPM="$(getPackageRoot)/rpm/$API_NAME"
ROCM_WHEEL_DIR="${BUILD_DIR}/_wheel"
BUILD_TYPE="Debug"
MAKE_OPTS="$DASH_JAY -C $BUILD_DIR"
SHARED_LIBS="ON"
CLEAN_OR_OUT=0;
MAKETARGET="deb"
PKGTYPE="deb"
WHEEL_PACKAGE=false
#parse the arguments
VALID_STR=$(getopt -o hcraso:p:w --long help,clean,release,static,address_sanitizer,outdir:,package:,wheel -- "$@")
eval set -- "$VALID_STR"
@@ -55,22 +51,22 @@ do
case "$1" in
-h | --help)
printUsage ; exit 0;;
-c | --clean)
TARGET="clean" ; ((CLEAN_OR_OUT|=1)) ; shift ;;
-r | --release)
BUILD_TYPE="Release" ; shift ;;
-a | --address_sanitizer)
set_asan_env_vars
set_address_sanitizer_on ; shift ;;
-s | --static)
SHARED_LIBS="OFF" ; shift ;;
-c | --clean)
TARGET="clean" ; ((CLEAN_OR_OUT|=1)) ; shift ;;
-o | --outdir)
TARGET="outdir"; PKGTYPE=$2 ; OUT_DIR_SPECIFIED=1 ; ((CLEAN_OR_OUT|=2)) ; shift 2 ;;
-p | --package)
MAKETARGET="$2" ; shift 2 ;;
-r | --release)
BUILD_TYPE="Release" ; shift ;;
-s | --static)
ack_and_skip_static ;;
-w | --wheel)
WHEEL_PACKAGE=true ; shift ;;
--) shift; break;; # end delimiter
--) shift; break;;
*)
echo " This should never come but just incase : UNEXPECTED ERROR Parm : [$1] ">&2 ; exit 20;;
esac
@@ -86,7 +82,6 @@ fi
clean() {
echo "Cleaning $PROJ_NAME"
rm -rf "$ROCM_WHEEL_DIR"
rm -rf "$BUILD_DIR"
rm -rf "$PACKAGE_DEB"
rm -rf "$PACKAGE_RPM"
@@ -97,10 +92,9 @@ clean() {
build() {
echo "Building $PROJ_NAME"
if [ "$DISTRO_ID" = centos-7 ]; then
echo "Skip make and uploading packages for Omniperf on Centos7 distro, due to python dependency"
echo "Skip make and uploading packages for ROCm Compute Profiler on Centos7 distro, due to python dependency"
exit 0
fi
if [ ! -d "$BUILD_DIR" ]; then
mkdir -p "$BUILD_DIR"
pushd "$BUILD_DIR" || exit
@@ -108,16 +102,16 @@ build() {
echo "ROCm CMake Params: $(rocm_cmake_params)"
echo "ROCm Common CMake Params: $(rocm_common_cmake_params)"
#install python deps
#python3 -m pip install -t ${BUILD_DIR}/python-libs -r ${ROCPROFILER_COMPUTE_ROOT}/requirements.txt
print_lib_type $SHARED_LIBS
cmake \
$(rocm_cmake_params) \
$(rocm_common_cmake_params) \
-DCHECK_PYTHON_DEPS=NO \
-DPYTHON_DEPS=${BUILD_DIR}/python-libs \
-DMOD_INSTALL_PATH=${BUILD_DIR}/modulefiles \
"$OMNIPERF_ROOT"
"$ROCPROFILER_COMPUTE_ROOT"
fi
make $MAKE_OPTS
make $MAKE_OPTS install
make $MAKE_OPTS package
@@ -126,22 +120,6 @@ build() {
copy_if RPM "${CPACKGEN:-"DEB;RPM"}" "$PACKAGE_RPM" "$BUILD_DIR/${API_NAME}"*.rpm
}
create_wheel_package() {
echo "Creating Omniperf wheel package"
# Copy the setup.py generator to build folder
mkdir -p "$ROCM_WHEEL_DIR"
cp -f "$SCRIPT_ROOT"/generate_setup_py.py "$ROCM_WHEEL_DIR"
cp -f "$SCRIPT_ROOT"/repackage_wheel.sh "$ROCM_WHEEL_DIR"
cd "$ROCM_WHEEL_DIR" || exit
# Currently only supports python3.6
./repackage_wheel.sh "$BUILD_DIR"/*.rpm python3.6
# Copy the wheel created to RPM folder which will be uploaded to artifactory
copy_if WHL "WHL" "$PACKAGE_RPM" "$ROCM_WHEEL_DIR"/dist/*.whl
}
print_output_directory() {
case ${PKGTYPE} in
("deb")
@@ -163,9 +141,4 @@ case "$TARGET" in
(*) die "Invalid target $TARGET" ;;
esac
if [[ $WHEEL_PACKAGE == true ]]; then
echo "Wheel Package build started !!!!"
create_wheel_package
fi
echo "Operation complete"
echo "Operation complete"

View File

@@ -37,7 +37,6 @@ PACKAGE_INCLUDE="$(getIncludePath)"
BUILD_DIR="$(getBuildPath $API_NAME)"
PACKAGE_DEB="$(getPackageRoot)/deb/$API_NAME"
PACKAGE_RPM="$(getPackageRoot)/rpm/$API_NAME"
ROCM_WHEEL_DIR="${BUILD_DIR}/_wheel"
PACKAGE_PREFIX="$ROCM_INSTALL_PATH"
BUILD_TYPE="Debug"
MAKE_OPTS="$DASH_JAY"
@@ -74,8 +73,7 @@ while true; do
shift
;;
-s | --static)
SHARED_LIBS="OFF"
shift
ack_and_skip_static
;;
-w | --wheel)
WHEEL_PACKAGE=true
@@ -113,7 +111,6 @@ fi
clean() {
echo "Cleaning $PROJ_NAME"
rm -rf "$ROCM_WHEEL_DIR"
rm -rf "$BUILD_DIR"
rm -rf "$PACKAGE_DEB"
rm -rf "$PACKAGE_RPM"
@@ -177,18 +174,6 @@ build_rocprofiler-sdk() {
fi
}
create_wheel_package() {
echo "Creating rocprofiler sdk wheel package"
mkdir -p "$ROCM_WHEEL_DIR"
cp -f "$SCRIPT_ROOT"/generate_setup_py.py "$ROCM_WHEEL_DIR"
cp -f "$SCRIPT_ROOT"/repackage_wheel.sh "$ROCM_WHEEL_DIR"
cd "$ROCM_WHEEL_DIR"
# Currently only supports python3.6
./repackage_wheel.sh "$BUILD_DIR"/*.rpm python3.6
# Copy the wheel created to RPM folder which will be uploaded to artifactory
copy_if WHL "WHL" "$PACKAGE_RPM" "$ROCM_WHEEL_DIR"/dist/*.whl
}
print_output_directory() {
case ${PKGTYPE} in
"deb")
@@ -214,9 +199,4 @@ case "$TARGET" in
*) die "Invalid target $TARGET" ;;
esac
if [[ $WHEEL_PACKAGE == true ]]; then
echo "Wheel Package build started !!!!"
create_wheel_package
fi
echo "Operation complete"

View File

@@ -14,7 +14,7 @@ printUsage() {
echo " -a, --address_sanitizer Enable address sanitizer"
echo " -o, --outdir <pkg_type> Print path of output directory containing packages of
type referred to by pkg_type"
echo " -w, --wheel Creates python wheel package of omnitrace.
echo " -w, --wheel Creates python wheel package of rocprof_sys.
It needs to be used along with -r option"
echo " -h, --help Prints this help"
echo
@@ -26,16 +26,18 @@ printUsage() {
return 0
}
API_NAME="omnitrace"
API_NAME="rocprofiler-systems"
PROJ_NAME="$API_NAME"
LIB_NAME="lib${API_NAME}"
TARGET="build"
MAKETARGET="deb"
PACKAGE_ROOT="$(getPackageRoot)"
PACKAGE_LIB="$(getLibPath)"
BUILD_DIR="$(getBuildPath $API_NAME)"
PACKAGE_DEB="$(getPackageRoot)/deb/$API_NAME"
PACKAGE_RPM="$(getPackageRoot)/rpm/$API_NAME"
BUILD_TYPE="Debug"
MAKE_OPTS="-j 8"
SHARED_LIBS="ON"
@@ -44,11 +46,11 @@ MAKETARGET="deb"
PKGTYPE="deb"
ASAN=0
#parse the arguments
VALID_STR=$(getopt -o hcraso:p:w --long help,clean,release,address_sanitizer,static,outdir:,package:,wheel -- "$@")
eval set -- "$VALID_STR"
while true; do
#echo "parocessing $1"
case "$1" in
-h | --help)
printUsage
@@ -65,17 +67,19 @@ while true; do
;;
-a | --address_sanitizer)
ack_and_ignore_asan
# set_asan_env_vars
# set_address_sanitizer_on
ASAN=1
shift
;;
-s | --static)
SHARED_LIBS="OFF"
shift
ack_and_skip_static
;;
-o | --outdir)
TARGET="outdir"
PKGTYPE=$2
# OUT_DIR_SPECIFIED=1
((CLEAN_OR_OUT |= 2))
shift 2
;;
@@ -84,13 +88,13 @@ while true; do
shift 2
;;
-w | --wheel)
echo "omnitrace: wheel build option accepted and ignored"
WHEEL_PACKAGE=true
shift
;;
--)
shift
break
;;
;; # end delimiter
*)
echo " This should never come but just incase : UNEXPECTED ERROR Parm : [$1] " >&2
exit 20
@@ -115,15 +119,11 @@ clean() {
rm -rf "$PACKAGE_LIB/${LIB_NAME:?}"*
}
build_omnitrace() {
build_rocprofiler_systems() {
echo "Building $PROJ_NAME"
if [ "$DISTRO_ID" = "mariner-2.0" ] || [ "$DISTRO_ID" = "ubuntu-24.04" ] || [ "$DISTRO_ID" = "azurelinux-3.0" ]; then
echo "Skip make and uploading packages for Omnitrace on \"${DISTRO_ID}\" distro"
exit 0
fi
if [ $ASAN == 1 ]; then
echo "Skip make and uploading packages for Omnitrace on ASAN build"
echo "Skip make and uploading packages for rocprofiler-systems on ASAN build"
exit 0
fi
if [ ! -d "$BUILD_DIR" ]; then
@@ -131,24 +131,69 @@ build_omnitrace() {
echo "Created build directory: $BUILD_DIR"
fi
cd $ROCPROFILER_SYSTEMS_ROOT || exit
echo "Current submodule status"
git submodule status
echo "Cached (old) submodule status"
git submodule status --cached
cat .git/config
echo "Updating submodules"
git submodule init
git submodule sync --recursive
git submodule update --init --recursive --force
echo "Updated submodule status"
git submodule status
cat .git/config
echo "Build directory: $BUILD_DIR"
pushd "$BUILD_DIR" || exit
print_lib_type $SHARED_LIBS
ELFUTIL_URL="https://compute-artifactory.amd.com/artifactory/rocm-generic-local/dev-tools/omnitrace/elfutils-0.188.tar.bz2"
BINUTIL_URL="https://compute-artifactory.amd.com/artifactory/rocm-generic-local/dev-tools/omnitrace/binutils-2.40.tar.gz"
echo "ROCm CMake Params: $(rocm_cmake_params)"
echo "ROCm Common CMake Params: $(rocm_common_cmake_params)"
echo "ELFUTIL_URL=$ELFUTIL_URL, BINUTIL_URL=$BINUTIL_URL"
if [ $ASAN == 1 ]; then
echo "Address Sanitizer path"
# Commenting out the below cmake command as it is not working as expected
# LD_LIBRARY_PATH=$ROCM_INSTALL_PATH/lib/asan:$LD_LIBRARY_PATH
# cmake \
# $(rocm_cmake_params) \
# $(rocm_common_cmake_params) \
# -DROCPROFSYS_BUILD_{LIBUNWIND,DYNINST}=ON \
# -DDYNINST_BUILD_{TBB,BOOST,ELFUTILS,LIBIBERTY}=ON \
# -DAMDDeviceLibs_DIR="${ROCM_INSTALL_PATH}/lib/asan/cmake/AMDDeviceLibs" \
# -Dhip_DIR="${ROCM_INSTALL_PATH}/lib/asan/cmake/hip" \
# -Dhip-lang_DIR="${ROCM_INSTALL_PATH}/lib/asan/cmake/hip-lang" \
# -Damd_comgr_DIR="${ROCM_INSTALL_PATH}/lib/asan/cmake/amd_comgr" \
# -Dhsa-runtime64_DIR="${ROCM_INSTALL_PATH}/lib/asan/cmake/hsa-runtime64" \
# -Dhsakmt_DIR="${ROCM_INSTALL_PATH}/lib/asan/cmake/hsakmt" \
# -DROCM_PATH="${ROCM_INSTALL_PATH}/lib/asan" \
# -Drocprofiler_ROOT_DIR="${ROCM_INSTALL_PATH}/lib/asan" \
# -DCMAKE_HIP_COMPILER_ROCM_ROOT="${ROCM_INSTALL_PATH}" \
# -DCMAKE_PREFIX_PATH="${ROCM_INSTALL_PATH};${ROCM_INSTALL_PATH}/lib/asan" \
# -DCMAKE_LIBRARY_PATH="${ROCM_INSTALL_PATH}/lib/asan" \
# -DCPACK_DEBIAN_PACKAGE_SHLIBDEPS=OFF \
# "$ROCPROFILER_SYSTEMS_ROOT"
else
cmake \
$(rocm_cmake_params) \
$(rocm_common_cmake_params) \
-DOMNITRACE_BUILD_{LIBUNWIND,DYNINST}=ON \
-DROCPROFSYS_BUILD_{LIBUNWIND,DYNINST}=ON \
-DDYNINST_BUILD_{TBB,BOOST,ELFUTILS,LIBIBERTY}=ON \
"$OMNITRACE_ROOT"
-DElfUtils_DOWNLOAD_URL="$ELFUTIL_URL" \
-D{DYNINST,TIMEMORY}_BINUTILS_DOWNLOAD_URL="$BINUTIL_URL" \
"$ROCPROFILER_SYSTEMS_ROOT"
fi
@@ -182,10 +227,10 @@ print_output_directory() {
verifyEnvSetup
case "$TARGET" in
clean) clean ;;
build) build_omnitrace ;;
outdir) print_output_directory ;;
*) die "Invalid target $TARGET" ;;
clean) clean ;;
build) build_rocprofiler_systems ;;
outdir) print_output_directory ;;
*) die "Invalid target $TARGET" ;;
esac
echo "Operation complete"

View File

@@ -8,7 +8,7 @@ printUsage() {
echo
echo "Options:"
echo " -c, --clean Clean output and delete all intermediate work"
echo " -s, --static Build static lib (.a). build instead of dynamic/shared(.so) "
echo " -s, --static Component/Build does not support static builds just accepting this param & ignore. No effect of the param on this build"
echo " -p, --package <type> Specify packaging format"
echo " -r, --release Make a release build instead of a debug build"
echo " -a, --address_sanitizer Enable address sanitizer"
@@ -42,9 +42,9 @@ SHARED_LIBS="ON"
CLEAN_OR_OUT=0
MAKETARGET="deb"
PKGTYPE="deb"
GPU_LIST="gfx900,gfx906,gfx908,gfx90a,gfx940,gfx941,gfx942,gfx1030,gfx1100,gfx1101,gfx1102"
GPU_LIST="gfx900,gfx906,gfx908,gfx90a,gfx940,gfx941,gfx942,gfx1030,gfx1031,gfx1100,gfx1101,gfx1102,gfx1200,gfx1201"
VALID_STR=$(getopt -o hcraso:p: --long help,clean,release,static,address_sanitizer,outdir:,package: -- "$@")
VALID_STR=$(getopt -o hcraswo:p: --long help,clean,release,static,wheel,address_sanitizer,outdir:,package: -- "$@")
eval set -- "$VALID_STR"
while true; do
@@ -68,7 +68,7 @@ while true; do
shift
;;
-s | --static)
SHARED_LIBS="OFF"
ack_and_skip_static
shift
;;
-o | --outdir)
@@ -131,7 +131,9 @@ build_rocprofiler() {
-DBUILD_SHARED_LIBS=$SHARED_LIBS \
-DENABLE_LDCONFIG=OFF \
-DUSE_PROF_API=1 \
-DUSE_GET_ROCM_PATH_API=1 \
-DGPU_TARGETS="$GPU_LIST" \
-DPython3_EXECUTABLE=$(which python3) \
-DPROF_API_HEADER_PATH="$WORK_ROOT/roctracer/inc/ext" \
-DHIP_HIPCC_FLAGS=$HIP_HIPCC_FLAGS";--offload-arch=$GPU_LIST" \
-DCPACK_OBJCOPY_EXECUTABLE="${ROCM_INSTALL_PATH}/llvm/bin/llvm-objcopy" \

Some files were not shown because too many files have changed in this diff Show More