76333 Коммитов
Автор SHA1 Сообщение Дата
Wenkai Du 5bebcb0015 Setup collectives threshold for enabling intranet (#387)
* Setup collectives threshold for enabling intranet

* Use separate operation counters for coll and p2p

[ROCm/rccl commit: b815a2800f]
2021-06-09 13:24:26 -07:00
Wenkai Du b815a2800f Setup collectives threshold for enabling intranet (#387)
* Setup collectives threshold for enabling intranet

* Use separate operation counters for coll and p2p
2021-06-09 13:24:26 -07:00
Bill(Shuzhou) Liu 673f5a4ee1 Disable bulk fetch. Add environment variable to enable it
RDC can optimize by bulk fetching multiple metrics using a single
rocm_smi call. However, currently this is not completely supported in
all ASIC generations. By default disable this for now.

Set environment variable RDC_BULK_FETCH_ENABLED=TRUE to enable
RDC bulk fetch.

BUG: SWDEV-289316

Change-Id: Ibb55514f198356dccf5f47bb0fd2d53c17acb251
2021-06-09 15:53:17 -04:00
Bill(Shuzhou) Liu 7d7a5bfd1c Disable bulk fetch. Add environment variable to enable it
RDC can optimize by bulk fetching multiple metrics using a single
rocm_smi call. However, currently this is not completely supported in
all ASIC generations. By default disable this for now.

Set environment variable RDC_BULK_FETCH_ENABLED=TRUE to enable
RDC bulk fetch.

BUG: SWDEV-289316

Change-Id: Ibb55514f198356dccf5f47bb0fd2d53c17acb251


[ROCm/rdc commit: 673f5a4ee1]
2021-06-09 15:53:17 -04:00
Wenkai Du ab9b9151d2 Add support for another Rome model (#385)
[ROCm/rccl commit: c2064adcc7]
2021-06-08 13:58:20 -07:00
Wenkai Du c2064adcc7 Add support for another Rome model (#385) 2021-06-08 13:58:20 -07:00
Stanley Tsang 52ffd67cd6 Updating changelog to show install script fix (#384)
[ROCm/rccl commit: 6842429a14]
2021-06-08 13:00:40 -06:00
Stanley Tsang 6842429a14 Updating changelog to show install script fix (#384) 2021-06-08 13:00:40 -06:00
rachida 4ba91a972c SWDEV-284863 Removed accidental override of recordid
recordid  cannot be just a counter. The code removed was doing
just that i.e. incrementing a counter. Recordid has to come
from recvals data structure. That code was left there since
a while when Evgeny and Rachida were trying to prototype this feature.
I am not sure why it was not spotted before.

Change-Id: Ia867066dcfca083fcd4111f2aefc2fec88c26314
2021-06-08 14:21:27 -04:00
rachida f2649f4049 SWDEV-284863 Removed accidental override of recordid
recordid  cannot be just a counter. The code removed was doing
just that i.e. incrementing a counter. Recordid has to come
from recvals data structure. That code was left there since
a while when Evgeny and Rachida were trying to prototype this feature.
I am not sure why it was not spotted before.

Change-Id: Ia867066dcfca083fcd4111f2aefc2fec88c26314


[ROCm/rocprofiler commit: 4ba91a972c]
2021-06-08 14:21:27 -04:00
Wenkai Du c8a432dc25 Allow intranode use of network connection (#383)
* Allow intranode use of network connection

* Checking for graph for null pointer

[ROCm/rccl commit: a3a8c2d56b]
2021-06-08 07:37:59 -07:00
Wenkai Du a3a8c2d56b Allow intranode use of network connection (#383)
* Allow intranode use of network connection

* Checking for graph for null pointer
2021-06-08 07:37:59 -07:00
Vladislav Sytchenko 261b237f17 SWDEV-289854 - Revert to using hardcoded library destination
Current packaging assumes that HIP runtime will always be installed in
/opt/rocm/lib. This is false to assume, because some distros like CentOS
will use the lib64 directory instead of lib. Relying on CMake to choose
the library directory for us will default in that case to lib64. Hence
there will be a mismatch between where HIP is installed and where CMake
thinks it is.

Change-Id: If6f881eab027bdbe8435d3de5295b00277e96a6a


[ROCm/hip commit: ed9e4ee7df]
2021-06-08 10:11:47 -04:00
Vladislav Sytchenko ed9e4ee7df SWDEV-289854 - Revert to using hardcoded library destination
Current packaging assumes that HIP runtime will always be installed in
/opt/rocm/lib. This is false to assume, because some distros like CentOS
will use the lib64 directory instead of lib. Relying on CMake to choose
the library directory for us will default in that case to lib64. Hence
there will be a mismatch between where HIP is installed and where CMake
thinks it is.

Change-Id: If6f881eab027bdbe8435d3de5295b00277e96a6a
2021-06-08 10:11:47 -04:00
Rahul Garg 50a4684ada ROCMOPS - 1956 - Update hipamd CMake
Change-Id: I1b6fa5953a27660b938d7232f0d88c5981624cd9


[ROCm/clr commit: e0d7fc8909]
2021-06-08 00:52:13 +00:00
Rahul Garg e0d7fc8909 ROCMOPS - 1956 - Update hipamd CMake
Change-Id: I1b6fa5953a27660b938d7232f0d88c5981624cd9
2021-06-08 00:52:13 +00:00
Sean Keely 981c6bd8c3 Remove unused GpuAgent.local_region_ member.
Change-Id: I99526e6b1f64e810f7fed5d922c540d252a46d80
2021-06-07 19:59:58 -04:00
Sean Keely 426a9f4df0 Remove unused GpuAgent.local_region_ member.
Change-Id: I99526e6b1f64e810f7fed5d922c540d252a46d80


[ROCm/ROCR-Runtime commit: 981c6bd8c3]
2021-06-07 19:59:58 -04:00
jujiang 2f5ad5b68d SWDEV-288929 - Add atomic system functions in HIP doc
Change-Id: Ic9ec276c1c486c4e1d3b83e1578081673f97efa2


[ROCm/hip commit: d8b2571d47]
2021-06-07 18:00:24 -04:00
jujiang d8b2571d47 SWDEV-288929 - Add atomic system functions in HIP doc
Change-Id: Ic9ec276c1c486c4e1d3b83e1578081673f97efa2
2021-06-07 18:00:24 -04:00
Rahul Garg 62a54a9ad5 ROCMOPS-1956 - Push restructured code to hipamd
hipamd will have AMD's ROCCLR based HIP backend implementation

Change-Id: Id7de9634519b4ce46fca71a1b61f3d5b1e3fc459


[ROCm/hipother commit: b2f4e62135]
2021-06-07 21:42:44 +00:00
Rahul Garg b2f4e62135 ROCMOPS-1956 - Push restructured code to hipamd
hipamd will have AMD's ROCCLR based HIP backend implementation

Change-Id: Id7de9634519b4ce46fca71a1b61f3d5b1e3fc459
2021-06-07 21:42:44 +00:00
Rahul Garg dd95d3d23c ROCMOPS-1956 - Push restructured code to hipamd
hipamd will have AMD's ROCCLR based HIP backend implementation

Change-Id: Id7de9634519b4ce46fca71a1b61f3d5b1e3fc459


[ROCm/clr commit: 19c84bc604]
2021-06-07 21:42:44 +00:00
Rahul Garg 19c84bc604 ROCMOPS-1956 - Push restructured code to hipamd
hipamd will have AMD's ROCCLR based HIP backend implementation

Change-Id: Id7de9634519b4ce46fca71a1b61f3d5b1e3fc459
2021-06-07 21:42:44 +00:00
Sean Keely bd59789f0b Add debugging checks for packet type in the scratch handler.
Change-Id: I84a6f18548ac39349595e3a1c8a5a9ff27d4e178
2021-06-07 15:36:18 -04:00
Sean Keely c35562e45b Add debugging checks for packet type in the scratch handler.
Change-Id: I84a6f18548ac39349595e3a1c8a5a9ff27d4e178


[ROCm/ROCR-Runtime commit: bd59789f0b]
2021-06-07 15:36:18 -04:00
Vladislav Sytchenko 43f722590c SWDEV-273235 - ocltst build fixes
Correctly reference the new amdocl target.

Do not use hardcoded paths to libamdocl64.so, instead query the path to
it from cmake.

Change-Id: I4738dd8d6d08f9fb72fdd725b6488b81342755a4


[ROCm/clr commit: 6046e6fea4]
2021-06-07 14:25:28 -04:00
Vladislav Sytchenko 6046e6fea4 SWDEV-273235 - ocltst build fixes
Correctly reference the new amdocl target.

Do not use hardcoded paths to libamdocl64.so, instead query the path to
it from cmake.

Change-Id: I4738dd8d6d08f9fb72fdd725b6488b81342755a4
2021-06-07 14:25:28 -04:00
Vladislav Sytchenko bb26e65554 SWDEV-289871 - Restore missing -Bsymbolic option
-Bsymbolic is required to make sure all AMD OpenCL runtime symbols are
resolved internally, otherwise ld might resolve them using symbols from
the OpenCL ICD. This will cause incorrect functions to be called.

Change-Id: I4ca0326a574ee2e537525bfda96ec20fabfe0158


[ROCm/clr commit: bc5c8df20b]
2021-06-07 14:22:53 -04:00
Vladislav Sytchenko bc5c8df20b SWDEV-289871 - Restore missing -Bsymbolic option
-Bsymbolic is required to make sure all AMD OpenCL runtime symbols are
resolved internally, otherwise ld might resolve them using symbols from
the OpenCL ICD. This will cause incorrect functions to be called.

Change-Id: I4ca0326a574ee2e537525bfda96ec20fabfe0158
2021-06-07 14:22:53 -04:00
cjatin 96abeb55fa SWDEV-245414 - Add Copyright and exlude mipmap test on Nvidia platform
Change-Id: Iba29caa0e9021488378625506197384c275a3282


[ROCm/hip commit: 575be36c2e]
2021-06-07 11:15:00 -04:00
cjatin 575be36c2e SWDEV-245414 - Add Copyright and exlude mipmap test on Nvidia platform
Change-Id: Iba29caa0e9021488378625506197384c275a3282
2021-06-07 11:15:00 -04:00
Yifan Zhang c24ed10dfa libhsakmt: add colon after KFDQMTest.SdmaConcurrentCopies
Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
Change-Id: Ief14e513e4b09db0607f5533a55f80d3b0be017e
2021-06-07 18:21:59 +08:00
Yifan Zhang b3e0b236df libhsakmt: add colon after KFDQMTest.SdmaConcurrentCopies
Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
Change-Id: Ief14e513e4b09db0607f5533a55f80d3b0be017e


[ROCm/ROCR-Runtime commit: c24ed10dfa]
2021-06-07 18:21:59 +08:00
Rahul Garg a5d7c0f92f SWDEV-288707 - Move amdhip.def
Change-Id: I77f9519d0d098b9cfdd1d09f67b3ab38d848620b


[ROCm/clr commit: f1f1d9d47c]
2021-06-07 05:16:16 +00:00
Rahul Garg f1f1d9d47c SWDEV-288707 - Move amdhip.def
Change-Id: I77f9519d0d098b9cfdd1d09f67b3ab38d848620b
2021-06-07 05:16:16 +00:00
Rahul Garg 5f17023409 SWDEV-288707 - Move amdhip.def
Change-Id: I77f9519d0d098b9cfdd1d09f67b3ab38d848620b


[ROCm/hip commit: 5d5e57ed60]
2021-06-07 05:16:16 +00:00
Rahul Garg 5d5e57ed60 SWDEV-288707 - Move amdhip.def
Change-Id: I77f9519d0d098b9cfdd1d09f67b3ab38d848620b
2021-06-07 05:16:16 +00:00
Yifan Zhang e72be0e54d kfdtest: Temporarily blacklist some svm related test cases for gfx902.
move blacklisted test case from gfx902 iommuv2 to dgpu path.

Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
Change-Id: I8b101226ca8dcd0c12c484f5f6ce12fe73a75bdc
Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
(cherry picked from commit 9cf4377572321396225950b9a58beb549120c2a3)
2021-06-06 23:07:29 -04:00
Yifan Zhang eb6d16072b kfdtest: Temporarily blacklist some svm related test cases for gfx902.
move blacklisted test case from gfx902 iommuv2 to dgpu path.

Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
Change-Id: I8b101226ca8dcd0c12c484f5f6ce12fe73a75bdc
Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
(cherry picked from commit 9cf4377572321396225950b9a58beb549120c2a3)


[ROCm/ROCR-Runtime commit: e72be0e54d]
2021-06-06 23:07:29 -04:00
Rahul Garg 7d8b728f8c SWDEV-288707 - Merge HIP refactored code to staging
These changes move-
- rocclr based implementation inside src/hipamd/src
- platform specific (both nvidia and amd) headers inside
src/hipamd/include/hip

Change-Id: Ia29791a727244952591fe1d813dcef0303b73a9e


[ROCm/hipother commit: 358b6fe5c4]
2021-06-07 00:07:24 +00:00
Rahul Garg 358b6fe5c4 SWDEV-288707 - Merge HIP refactored code to staging
These changes move-
- rocclr based implementation inside src/hipamd/src
- platform specific (both nvidia and amd) headers inside
src/hipamd/include/hip

Change-Id: Ia29791a727244952591fe1d813dcef0303b73a9e
2021-06-07 00:07:24 +00:00
Rahul Garg 28f8d08148 SWDEV-288707 - Merge HIP refactored code to staging
These changes move-
- rocclr based implementation inside src/hipamd/src
- platform specific (both nvidia and amd) headers inside
src/hipamd/include/hip

Change-Id: Ia29791a727244952591fe1d813dcef0303b73a9e


[ROCm/clr commit: 260a1fed06]
2021-06-07 00:07:24 +00:00
Rahul Garg 260a1fed06 SWDEV-288707 - Merge HIP refactored code to staging
These changes move-
- rocclr based implementation inside src/hipamd/src
- platform specific (both nvidia and amd) headers inside
src/hipamd/include/hip

Change-Id: Ia29791a727244952591fe1d813dcef0303b73a9e
2021-06-07 00:07:24 +00:00
Rahul Garg 4cd9afebe9 SWDEV-288707 - Merge HIP refactored code to staging
These changes move-
- rocclr based implementation inside src/hipamd/src
- platform specific (both nvidia and amd) headers inside
src/hipamd/include/hip

Change-Id: Ia29791a727244952591fe1d813dcef0303b73a9e


[ROCm/hip commit: ce76ec8b88]
2021-06-07 00:07:24 +00:00
Rahul Garg ce76ec8b88 SWDEV-288707 - Merge HIP refactored code to staging
These changes move-
- rocclr based implementation inside src/hipamd/src
- platform specific (both nvidia and amd) headers inside
src/hipamd/include/hip

Change-Id: Ia29791a727244952591fe1d813dcef0303b73a9e
2021-06-07 00:07:24 +00:00
Vladislav Sytchenko ca460b03ee SWDEV-273235 - Enable building clinfo/cltrace on Windows
This patch also includes additional cleanup logic to the installation
rules. The OpenCL ICD already defines its own installation rules, no
need to duplicate them.

Change-Id: Id1c386307a83b82e21f6fa132e41957d9b348fd3


[ROCm/clr commit: c6eadce180]
2021-06-05 12:52:31 -04:00
Vladislav Sytchenko c6eadce180 SWDEV-273235 - Enable building clinfo/cltrace on Windows
This patch also includes additional cleanup logic to the installation
rules. The OpenCL ICD already defines its own installation rules, no
need to duplicate them.

Change-Id: Id1c386307a83b82e21f6fa132e41957d9b348fd3
2021-06-05 12:52:31 -04:00
Greg Inozemtsev 45c28c6c36 Cleanup argument error handling and messages
Add error checking for minbytes and maxbytes arguments

Also accept lowercase literals when parsing size arguments and print errors and usage on stderr.


[ROCm/rccl-tests commit: c4de829d91]
2021-06-04 21:47:40 +00:00
Greg Inozemtsev c4de829d91 Cleanup argument error handling and messages
Add error checking for minbytes and maxbytes arguments

Also accept lowercase literals when parsing size arguments and print errors and usage on stderr.
2021-06-04 21:47:40 +00:00