Commit graph

12343 Commits

Autor SHA1 Nachricht Datum
Matt Arsenault 4af8b4ef0f SWDEV-1 - Fix -Winconsistent-missing-override
Change-Id: Ie69f234c755b112b7fe8ef7d00b8720bc178044c


[ROCm/clr commit: 98e0607f0e]
2023-06-23 09:04:28 -04:00
Ioannis Assiouras 320d7c1d9a SWDEV-405182 - Revert min alignment for abstract parameters stack to 16 bytes
Change-Id: I9e6ace281468e8ef11b011c58f5971ce8907f3c6


[ROCm/clr commit: 4add0e6563]
2023-06-23 04:39:51 -04:00
Rahul Garg 9293544234 SWDEV-389099 - Add initial CHANGELOG
Change-Id: I9091860684345a647ebe7167eb22a3aaf8df69a6


[ROCm/clr commit: 53ad3c6e5d]
2023-06-23 02:35:12 -04:00
sdashmiz 6c174f79fa SWDEV-405485 - move the param validation
Signed-off-by: sdashmiz <shadi.dashmiz@amd.com>
Change-Id: Ic3a27c47a88954da866a91494bcfb8721f33ad2b


[ROCm/clr commit: 2cdaf7e80f]
2023-06-22 23:33:48 -04:00
Alex Xie 867d1fc66f SWDEV-382072 - One more case that needs workaround
Change-Id: Ief81dd898e4d59ba858f38a843d79be45b5ea944


[ROCm/clr commit: cfc10d0e1e]
2023-06-22 19:00:34 -04:00
taosang2 9aa3a87b66 SWDEV-368553 – Add missing mipmap Apis
Add missing mipmap Apis’ implementation.
Fix some bugs of mimpmap apis.
Use hipmipmappedArray to differentiate cuda
and driver apis on Nvidia.
Change-Id: I6079d9f3b2ddf4e42b9a6f7f3902322cfca02cfd


[ROCm/clr commit: f03c11491b]
2023-06-22 17:24:21 -04:00
Alex Xie c14a017d77 SWDEV-382072 - [Naiv 10]EncodeSDK H264 test fails
let ROCr and addrlib decided whether a pitch is supported.
Then, check if we uses workaround.

Also improve image performance for Navi 1x in some cases.

Change-Id: If8bda46648b23c1dcfa26608ca3bb9cd89fef430


[ROCm/clr commit: b961d4a970]
2023-06-22 10:40:31 -04:00
pghafari b01ca8b9c7 SWDEV-406821 - use int32 for grid y,z till rocm 6
Change-Id: If9eee4dd5553cc81e283cb2211cc69884808d5c1


[ROCm/clr commit: 9e08ecdd16]
2023-06-22 09:42:00 -04:00
Ioannis Assiouras a1c01b7d34 SWDEV-406894 - Fix segfault when access HIPRTC_JIT_LOG_VERBOSE from options buffer
Change-Id: I24143dfb726d9a71fdb17b03d5b5dd910b54f5bb


[ROCm/clr commit: 0af9788e32]
2023-06-22 00:53:29 -04:00
Sourabh Betigeri c00cd118d3 SWDEV-402274 - Adds CG API group_dim() in class thread_block
Change-Id: I1c2569fffbc6fc730df9ba80786459c71140d327


[ROCm/clr commit: 422d3ea035]
2023-06-21 20:10:19 -04:00
German c8e65415ea SWDEV-311270 - Clean-up IPC path in ROCclr
- Make IPC interfaces generic between devices and rely on the IPC buffer
for attach/detach logic

Change-Id: Id3c18d122030329b7ee532bbb6317de9dd6a0bbe


[ROCm/clr commit: 9aa6f25f1c]
2023-06-20 15:18:21 -04:00
Saleel Kudchadker 60b9dc4a45 SWDEV-301667 - Better log
Change-Id: I3b5b3593cb0144e56a4ddea2fa29028544a73203


[ROCm/clr commit: b2240bcda5]
2023-06-20 14:26:44 -04:00
Sourabh Betigeri 395c4f1b40 SWDEV-406750 - Fix to avoid a warning when compiling HIP runtime
Change-Id: If1d848d334bc0121a96655781fc60b0321213398


[ROCm/clr commit: 0330801846]
2023-06-20 12:16:16 -04:00
Konstantin Zhuravlyov cdaea3b1be SWDEV-325538 - Enable code object v5 by default
Change-Id: I8c2bd415461403366de6e8b9157dd2fa6dc0c2a6


[ROCm/clr commit: 5914ac3c6e]
2023-06-20 11:08:50 -04:00
German 00b9c77a3a SWDEV-404940 - Pass -cl-uniform-work-group-size to clang
Change-Id: I460dd848d439204e60f618669aef86f77e094c1a


[ROCm/clr commit: f13988b88b]
2023-06-19 19:44:29 -04:00
Jeremy Newton 8b0ba068d3 Install .hipVersion into datadir for linux
Starting with LLVM 16, we can put .hipVersion in datadir for better
Linux FHS compliance:
https://github.com/llvm/llvm-project/commit/082593ff7aff68060bd66dccfa43493d07d9c255

For Windows, we can keep it status quo.

Change-Id: Id15ed0dac346fabec21ee060e7576ce085b1af70


[ROCm/clr commit: 03bfa61684]
2023-06-19 11:41:46 -04:00
Saleel Kudchadker 2e67d925a1 SWDEV-384557 - Include query under env var
Change-Id: Ibe0335168f9ad0cb72abb2ed69a5e40d33ce7c1c


[ROCm/clr commit: 4510caed26]
2023-06-16 15:09:46 -07:00
ajay c4e5952687 SWDEV-404940 - __amd_rocclr_initHeap in pal.
Avoid Seg fault on Windows

Change-Id: I3901c003dd33222ec9e2600c519d8503f5e81c01


[ROCm/clr commit: 169d17f47f]
2023-06-16 11:48:42 -04:00
Jatin Chaudhary 0cd45b8cc7 SWDEV-399522 - Fix the issue #2 raised on github clr repo. Adds signed long long atomics
Change-Id: I09c386d45e219e472d2da1449f678b9d17fe3e2e


[ROCm/clr commit: be02e1e524]
2023-06-16 08:53:45 -04:00
Jaydeep Patel de6662f1b7 SWDEV-403768 - Check mem type is managed and result mem addresses are not null.
Change-Id: Ib40c468a981e0dd42b2da65aeb7543b11af4bc90


[ROCm/clr commit: a6e72de83d]
2023-06-16 08:49:21 -04:00
Jaydeep Patel 600ad0e9dc SWDEV-403770 - Add validations.
Change-Id: If36202ba9777d2a0d6f8ee92982dbff92bafe81a


[ROCm/clr commit: 0a20da39cd]
2023-06-16 02:37:25 -04:00
sdashmiz 2216908962 SWDEV-403638 - Fix warnings
- disable deprecated function use warning
- disalbe size_t to .type' warning
- disable conversion from 'type1' to 'type2' warning

Signed-off-by: sdashmiz <shadi.dashmiz@amd.com>
Change-Id: I64161fd37cf56de3d132102103267ae8da40193a


[ROCm/clr commit: 38a67df312]
2023-06-15 12:17:22 -04:00
sdashmiz 2e960fae8f SWDEV-403063 - Clean up hip graph
- Replace dynamic allocation with embedded struct in hip graph

Signed-off-by: sdashmiz <shadi.dashmiz@amd.com>
Change-Id: Idfe3f7393fa50d6e510d4c88f779408f96d3accb


[ROCm/clr commit: cc85289739]
2023-06-15 12:17:09 -04:00
German af5944dc71 SWDEV-311270 - Add IPC support for memory pools
Initial implementation for hipMemPoolExportToShareableHandle,
hipMemPoolImportFromShareableHandle,
hipMemPoolExportPointer and hipMemPoolImportPointer

Change-Id: I0ebdc48e9163b394ded560adca6c38bbc5aee7d1


[ROCm/clr commit: 1a0c3e4dc4]
2023-06-15 11:36:52 -04:00
Jaydeep Patel d6086f9d69 SWDEV-403766 - Release mem obj while device reset.
Change-Id: I8a93061c9c4ed692bf7ba1b872af370f79a13a5d


[ROCm/clr commit: ec75a3e60e]
2023-06-15 09:52:50 -04:00
Jatin Chaudhary 8913c4de4f SWDEV-400136 - change make_complex function name to make_hipComplex
Change-Id: Ia276610286df80ec35fe11f303eb9919eeda87dd


[ROCm/clr commit: f7f70a0880]
2023-06-15 09:02:38 -04:00
Saleel Kudchadker 87652b6c93 SWDEV-384557 - Use toggle for SDMA query
- Use HIP_USE_SDMA_QUERY env var toggle for new API use. Env var is 0 by
default

Change-Id: If725a0c41e15f78a1a6c3f47942954fe9240b4db


[ROCm/clr commit: 8d193c32bb]
2023-06-15 01:02:24 -04:00
Jacob Lambert 52eaffa1a4 Revert "SWDEV-371628 - Shift device lib linking into clang driver"
This reverts commit 2e664d2492.

Reason for revert: Performance regressions and failures observed. Need to investigate those and before re-applying patch

Change-Id: I42ba0605797f9bdcfb5d5102927dd01405cf05e3


[ROCm/clr commit: 8047d8e3e8]
2023-06-15 00:56:08 -04:00
Jason Tang 6edd795788 SWDEV-404928 - Remove rocm-opencl depends of rocm-ocl-icd
Will replace rocm-ocl-icd with packaging of khronos opencl-icd-loader

Change-Id: I26da58e80fd1426b9908d0b524f93b17ab27626d


[ROCm/clr commit: a4bba6d705]
2023-06-14 14:35:57 -04:00
Ioannis Assiouras af03b2c8d2 SWDEV-386069 - Added nvidia and AMD config.cmake files as part of the dev package
Change-Id: I3c9a65b287822d91407996ca59ac90051b749923


[ROCm/clr commit: d2e36bc047]
2023-06-07 04:55:52 -04:00
Jaydeep Patel 063b3d4e51 SWDEV-403995 - uniform_work_group_size is interger type is 1 for true and 0 for false. Lighting kernel's init sets uniform work group flag and hence not need to set it from options.
Change-Id: Ife45c1e3ccef03d16d7e07c3e0714fd8fd388fd3


[ROCm/clr commit: 2c9c58e7bb]
2023-06-06 23:05:25 -04:00
Rahul Garg cea26614db SWDEV-379678 - Take hipcc/hipconfig from hipcc
Change-Id: I53a6e5108e142860747186c0e7ed4fc2cb384a69


[ROCm/clr commit: 17c052ca9a]
2023-06-06 17:21:59 -04:00
jatang 34f68be605 SWDEV-373396 - OCLCreateBuffer: make the CPU checkResult() 4 times faster
Change-Id: If20cd6b509896a748f74ab5296cb85f2c4a9f04c


[ROCm/clr commit: a456182888]
2023-06-06 17:15:34 -04:00
Ioannis Assiouras 917dc7ee82 SWDEV-401852 - Do not define hipCreateChannelDesc for long types in LP64
Change-Id: I56881a229ea666269d575c6bdfde5081562ec009


[ROCm/clr commit: 4f5dfa77bd]
2023-06-06 14:28:15 -04:00
Jaydeep Patel 7be6ca6306 SWDEV-404118 - Change to uint.
Change-Id: Ib9ced171a4330e868a584bfb86d7ee5418e784ed


[ROCm/clr commit: 9daf8e3177]
2023-06-06 13:17:56 -04:00
Ajay 06213e3b4b SWDEV-400515 - populate __HIP_NO_IMAGE_SUPPORT flag
MI300 does not support image APIs.
Apps to use __HIP_NO_IMAGE_SUPPORT instead of arch

Change-Id: I91178bfd27ea7b7188e7a958a876c0264f4469aa


[ROCm/clr commit: 16c6b365c2]
2023-06-06 12:11:41 -04:00
German 17c6b36204 SWDEV-403871 - Add MS platform recognition
Change-Id: Ie01e876f9322e35f4430352d1e4328f613133ee5


[ROCm/clr commit: c0e9734c35]
2023-06-06 11:57:07 -04:00
Maneesh Gupta 0be27c38f8 SWDEV-397358 - Revert "Return hipErrorNotSupported for unsupported APIs"
This reverts commit 32a73102ea.

Change-Id: Idb18e1ce6bcfc1da144fabf1ba01329f4f6604a5
(cherry picked from commit 5c38999efe11243dcdd8cc5244172e6840c47492)


[ROCm/clr commit: 68238c99d2]
2023-06-05 04:55:33 -04:00
Satyanvesh Dittakavi d7b529ff6e SWDEV-395996 - Add double precision constants to math constants header
Fixes a typo and adds missing constants matching cuda

Change-Id: Iebf47d107af361cbd356c2a3b9cd6ef121d01363


[ROCm/clr commit: 287f607d2b]
2023-06-05 03:31:27 -04:00
Jaydeep Patel 911002f262 SWDEV-403762 - Fill pciBusId even if len <=12.
Change-Id: Iaf1b3314f4ae5597d42d54eb4c977667ecbb9cfb


[ROCm/clr commit: 0064fac887]
2023-06-02 10:47:21 +00:00
Ranjith Ramakrishnan fe2a510c08 SWDEV-366831 - Compile time flag to switch between #warning and #error message
Using backward compatibility paths will provide #warning message be default.
Compile time option added to enable/disable the #error message.
Updated the backward compatibility message

Change-Id: I3bab00df26145991b32176d2d76977c2e953bf5f


[ROCm/clr commit: f788150132]
2023-06-01 04:22:19 -04:00
Anusha GodavarthySurya 729779a3b3 SWDEV-399098 - Change hip_init lock to use std::call_once
Updating hip_init lock to use std::call_once fixed Performance
drops in TF benchmarks for FP16

Change-Id: Ib1125ac66806b232057ba183e296ea4d0642d08d


[ROCm/clr commit: 2f83719d12]
2023-06-01 02:04:29 -04:00
Jacob Lambert 7ee73c61b0 SWDEV-375055 - Re-enable Comgr unbundler
With recent upstream changes (D145770), we can now use the
Comgr unbundler without requiring an env field in the supplied
targetID. For users, this is consistent with previous legacy
unbundler behavior.

Change-Id: I5f085b0fa1ad352bbbb282b75367c206b75f279f


[ROCm/clr commit: 443f912c7f]
2023-05-31 16:14:08 -04:00
Jacob Lambert 911dbabdff SWDEV-380857 - Remove space in "Linker Program" used to set Comgr object name
Not a required change, but it does make dealing with temporary files generated
by Comgr easier.

Change-Id: I9c43138dd2a6c4fea965b57fbce7a087ab2bbd28


[ROCm/clr commit: 1171518b97]
2023-05-30 13:47:26 -04:00
Jaydeep Patel 6ab03957bb SWDEV-401847 - Update atomicMin/Max for float and double.
Change-Id: Ib5be459b8a24f0739e299ed12c9f877f8baa02b1


[ROCm/clr commit: d576356505]
2023-05-29 14:27:16 -04:00
sdashmiz 9a3d3aef8c SWDEV-360031 - Event record clean up
Signed-off-by: sdashmiz <shadi.dashmiz@amd.com>
Change-Id: Ic7b5edd501c5f61b2bce87252ac040cbc4083006


[ROCm/clr commit: 1e9dfdfe99]
2023-05-29 11:42:16 -04:00
Vikram 5ba0fb5802 SWDEV-354050 - Runtime support for HIP non hostcall printf
Relates to https://reviews.llvm.org/D150427,

Each printf call populates buffer with following data
1. Control DWord - contains info regarding stream, format string constness and size of data frame
   (see http://gerrit-git.amd.com/c/lightning/ec/device-libs/+/857722 for more info)
2. Hash of the format string (if constant) else the format string itself
3. Printf arguments (each aligned to 8 byte boundary)

Change-Id: I7e320deb343921b4b4cfaf08a2be2883e0bc1f65


[ROCm/clr commit: 7b6a8f1702]
2023-05-26 01:55:20 -04:00
Jeremy Newton b1266e557e SWDEV-361988 - Fix hip library install permissions
"FILES" installs files as 644, but we want libraries to be 755, which
we can do with "PROGRAMS".

Signed-off-by: Jeremy Newton <Jeremy.Newton@amd.com>
Change-Id: I155ed77482839ff6d71f90239a014d239e20f4b8


[ROCm/clr commit: 2cda949920]
2023-05-25 22:47:32 -04:00
Satyanvesh Dittakavi 094b2bfa9e SWDEV-401850 - Fix hipCreateChannelDescHalf APIs
- Address ChannelDescHalf1 and ChannelDescHalf2 APIs not returning
correct Channel Descriptors in amd headers
- Add missing hipCreateChannelDescHalf APIs in nvidia headers

Change-Id: I558847425c2459d3cde329ca2e926d882bb0a9dd


[ROCm/clr commit: 586731c2cc]
2023-05-25 04:36:24 -04:00
Ioannis Assiouras babff0ff73 SWDEV-401902 - Set default CMAKE_BUILD_TYPE in the top level CMakeLists.txt
Change-Id: I3a3e24da2dc7eec5feb1f748ea49af3df40d4cd9


[ROCm/clr commit: f945adf0bd]
2023-05-24 14:11:18 -04:00