AI/rocm-systems
2
0
Bifurcation 0
Code Tickets Demandes d'ajout Actions Paquets Projets Publications Wiki Activité
76 333 Révisions 3 Branches 0 Étiquette
develop
Graphe des révisions

1338 Révisions

Auteur SHA1 Message Date
Juan Manuel MARTINEZ CAAMAÑO 16a5bfd654 SWDEV-287018 - Remove requiredDump argument from Program::linkLLVMBitcode 
Weirdly, the `requiredDump` argument to linkLLVMBitcode was used to enable/disable
the keeping temporary bytecode files (those generated by -save-temps=all) after linking.

This patch removes this argument as there is no obvious benefit from keepeing it
(the user would only rely on -save-temps=all to control this).

Change-Id: I0c00486f95eb1d4e296b5247c488407c47f0b2d9


[ROCm/clr commit: 8ab3fd58cf]
 2023-01-16 03:34:21 -05:00
Anusha GodavarthySurya e8c8fd8c3f SWDEV-374778 - Select SDMA path when HSA_XNACK is enabled 
Change-Id: I052675186238eb24b910a0a558d6b8ede36d1413


[ROCm/clr commit: 710749c291]
 2023-01-12 23:34:21 -05:00
Ioannis Assiouras ab81da6544 SWDEV-374401 - Surface the ammount of used constant memory by kernel 
Change-Id: I6de0e46482f27b9068582fdaee0e562f9f71f9f0


[ROCm/clr commit: 4d49204fb9]
 2023-01-12 05:22:37 -05:00
Saleel Kudchadker f903a6c2c9 SWDEV-371123 - Fix barrier value packet header 
- Use correct header for vendor packet
- Pass one dependent signal when submitting a marker if there is one

Change-Id: I4efc70dd5204b559de26f899d0637f50421c8834


[ROCm/clr commit: 77ceb7e9ad]
 2023-01-12 04:06:03 -05:00
jatang 1f5f1c8708 SWDEV-287018 - Only generate _optimized.bc when both requiredDump and DUMP_BC_LINKED are true 
The current code generates a _optimized.bc regardlessly, so put back the original logic.

Change-Id: I3f84d10934b3e983f5f828af8d0943449a6e1d94


[ROCm/clr commit: 6647882773]
 2023-01-11 17:51:35 -05:00
Anusha GodavarthySurya 379781ea65 SWDEV-364576 - initialize device malloc heap state using blit kernel 
Change-Id: I5d0172aff7d2c04b322a4d828b8a2b438158b80f


[ROCm/clr commit: 274f2de391]
 2023-01-07 06:53:53 +00:00
Saleel Kudchadker 4a92208bb7 SWDEV-371123 - Fix typo for log print 
Change-Id: I18297a032d6fc9a030863eac7c525e86e9fdcd29


[ROCm/clr commit: 6651aab28d]
 2023-01-05 17:46:22 -05:00
jatang fca7b357b8 SWDEV-372791 - Default UniformWorkGroupSize to true for CL1.x 
Change-Id: If03396b5a7a3a69e05fdb4c43791a0d84d9f4127


[ROCm/clr commit: 05a8eaaf0f]
 2022-12-28 12:59:08 -05:00
Jaydeep Patel f1eae835e3 SWDEV-374370 - Propogate element size to blit kernel. 
Change-Id: I06d1ae6feebd238e9a63c617eb4c4dcf485d9ee0


[ROCm/clr commit: 070ae4e6d4]
 2022-12-26 09:33:50 +00:00
Jaydeep Patel 7df45ffc38 SWDEV-366087 - Pass pitch and slice pitch to blit kernel from rect struct. 
Change-Id: I1ffe54929db59a40e2a1ae19c125f8d8e81b07ec


[ROCm/clr commit: 9076d9a518]
 2022-12-20 16:43:49 +00:00
Saleel Kudchadker ed4c04b150 SWDEV-373334 - Use copyMetadata for blit decisions 
- Check isAsync flag for small host copies on large bar as it synchronizes
- Use CopyEngine Preference hint if HMM is enabled.

Change-Id: I1ffc4b2604ed03cf5979cdc454178648c5ae5cba


[ROCm/clr commit: e0384f9f6b]
 2022-12-15 17:09:02 -05:00
Jacob Lambert 87e6e1357b SWDEV-356171 - Adding new Comgr mangled name fetching APIs 
amd::Comgr::populate_mangled_names()
  amd::Comgr::get_mangled_name()

Change-Id: Ie6ebf3aae76632f7e7cfef3c90642b2e3cd10b48


[ROCm/clr commit: 50d77d17fd]
 2022-12-14 12:40:18 -05:00
jatang b9cb9b64ed SWDEV-1 - Fix another build warning. 
Change-Id: Ifed33732752e9d9ec74a244a0104bf571a9fc6c5


[ROCm/clr commit: a4fadb191a]
 2022-12-14 10:14:49 -05:00
jatang 3707f5d671 SWDEV-1 - Not to truncate size_t in print when it's 64bit 
Change-Id: Iab8b7eb128c767b6808b19cbeac5169b7a97f1ad


[ROCm/clr commit: e8cd1e93e8]
 2022-12-13 09:13:33 -05:00
kjayapra-amd e295beb8ba SWDEV-371904 - Adding pseudo fine grain flag to hsa memory allocation for device fine grained memory. 
Change-Id: I8cada90f0e3880dfbc5bf5a3fac4554e7a0cb08e


[ROCm/clr commit: e56a611b92]
 2022-12-11 08:15:17 -05:00
Todd tiantuo Li bf1d819ac7 SWDEV-354868 - Queue::Create() for RT queue should fail when number of reserved CU's for RT queue is zero (most likely due to being aligned down by dedicatedCuGranularity). 
Change-Id: I234e7ff83cb312bf44f5ad4b1a897c079f5106a9


[ROCm/clr commit: 9168415ca2]
 2022-12-09 16:45:10 -05:00
Ioannis Assiouras 733c8d1d1c SWDEV-369581 - Convey copy API metadata to ROCclr 
Change-Id: I569462d6d268700d419510255e201bf7d80d6714


[ROCm/clr commit: 72b45e2a1f]
 2022-12-09 00:27:15 -05:00
pghafari 4e6a699eae SWDEV-366279 - updating p2p error msg 
Change-Id: I294e7770efd11f511dd5817eb28bd3c97d9d4926


[ROCm/clr commit: 402504f548]
 2022-12-09 00:05:04 -05:00
Jaydeep Patel be5d6fb005 SWDEV-370537 - PAL does 64kb alignment so no need to do it in runtime. Reverting alignment to 4kb. 
Change-Id: If3eaac65bf63bb9b8b73a3cad1bb34d357f1adeb


[ROCm/clr commit: cd7034c9bb]
 2022-12-08 14:54:46 -05:00
German ebf7fad258 SWDEV-370901 - Force sysmem allocation 
Adobe AE allocates/destroys system memory per each frame
in some effects, which is a very expensive operation.
Disable pinning and force sysmem allocation in VidMM.
That allows to trigger resource cache in runtime and
avoid paging operations, caused by pinning.
The actual benchmark doesn't have significant improvement
from the change, but GPU view traces are much cleaner.

Change-Id: I92cdd87d37c64611788f05250f71d6fe489b4d7e


[ROCm/clr commit: 8cf53e7dff]
 2022-12-08 13:06:51 -05:00
Saleel Kudchadker 7ba49616e9 SWDEV-371123 - Use barrier value packet for event records 
Change-Id: I5e5e5e89e0d96a2430b4682d168b76848fa5b94e


[ROCm/clr commit: 4f64d89026]
 2022-12-07 17:57:36 -05:00
Tao Sang 0efc782e69 SWDEV-347733 - Remove luxmark workaround for gfx11 
Make gfx11 always use LC by default.
Fix switch order of Navi3x.
Change-Id: I5774563edf4d1aed29a4e92723630adbfb75c48c


[ROCm/clr commit: 8c582a02ef]
 2022-12-05 18:38:08 -05:00
Sourabh Betigeri 7aa958a8f7 SWDEV-305894 - Cooperative groups grid and multi grid sync support for gfx940+ 
Change-Id: I35d72f1cb50c3a96eee56a612b72d641852b145f


[ROCm/clr commit: 5d7f3f9f3c]
 2022-12-05 16:30:30 -05:00
Saleel Kudchadker f14b550e87 SWDEV-371123 - Refactor dispatchBarrierValuePacket 
Change-Id: I846bd3f60dd8db125e2ca5475e9d25fdd567922b


[ROCm/clr commit: 10ecf8f159]
 2022-12-02 12:22:31 -05:00
Jason Tang 5cc5cf4f11 SWDEV-1 - Fix a build error. 
Change-Id: I5f14cdca04ee9867f0ae63fff197a0f21ee5f1d2


[ROCm/clr commit: 56fe793134]
 2022-12-01 10:31:34 -05:00
Jason Tang 310f43e9e3 SWDEV-1 - Fix a build warning. 
Change-Id: I085fcc68ef3aaebf85515ea45fca99d7ec845721


[ROCm/clr commit: c825d06454]
 2022-11-30 11:07:15 -05:00
Ioannis Assiouras 941dc51bfd SWDEV-369547 - Ensure that ipc_mem_detach is not called for non-ipcShared MemObj. 
Change-Id: Ia73b60b9f08d593301879e0f72c395edbc215112


[ROCm/clr commit: b445717f72]
 2022-11-30 02:16:13 -05:00
kjayapra-amd 1fea133877 SWDEV-358915 - Wait for 10 micro secs, if the next engine is SDMA. 
Change-Id: Ib3ed91510e62090f47d4f55be8de635caaee083d


[ROCm/clr commit: 61b6a94627]
 2022-11-28 12:14:16 -05:00
Anusha GodavarthySurya a93695d9f0 SWDEV-354074 - cache hsa queue and delete when device is destroyed 
Change-Id: I26365521d785f0bc612e32bdcdb6caacb3af9a11


[ROCm/clr commit: 446a3fc688]
 2022-11-23 07:17:45 +00:00
Alex Xie 1ac90df635 SWDEV-363271 - Add some performance counter logs 
Change-Id: I368ba4afe47357eb6051b2a06f762b8050e5a560


[ROCm/clr commit: 9e00c674c8]
 2022-11-19 16:27:14 -05:00
German e7d9a3f450 SWDEV-368308 - Remove HW debug extension 
Change-Id: If0c68023c09f0dac9111d52ecc0ad63719aa4e70


[ROCm/clr commit: e5a36ab1ad]
 2022-11-18 10:29:44 -05:00
German 4b6a6ba8e8 SWDEV-363074 - Adjust staging copy limits in Windows 
Pinned copy can cause big performance drops, because slow pinning under Windows.
Use up to 128MB for staging transfers. Change staging buffer size to 4MB.
Linux path should still have the old defaults.

Change-Id: I954edceb3ec89e8e670be116aa2d0a9564c8b11c


[ROCm/clr commit: 79d12df147]
 2022-11-17 14:48:16 -05:00
German cb4aa60a37 SWDEV-328670 - Enable ROCr interop 
Change-Id: I932afa2a0e75b1f1771c4be1e984b32966636afb


[ROCm/clr commit: b40823ccc1]
 2022-11-17 09:40:34 -05:00
Juan Manuel MARTINEZ CAAMAÑO ed84e3c9fa SWDEV-287018 - Generate tempraries for _optimized.bc and _linked.bc if -save-temp-all is passed. 
Change-Id: Ie8a7af66f0764f0a94948d45d7deed786f799c96


[ROCm/clr commit: 9b8005d758]
 2022-11-15 07:49:07 -05:00
Jason Tang 4eb6a108a3 SWDEV-1 - Add LOG_ERROR output for Pal::CreatePlatform failure 
Change-Id: Id5c3fdcd8f822bfa67ae570c09ede7ddc8a936ef


[ROCm/clr commit: 2c09557d60]
 2022-11-10 09:30:10 -05:00
Maneesh Gupta f04740b919 SWDEV-299940 - Query global memory available on device with HSA attribute HSA_AMD_AGENT_INFO_MEMORY_AVAIL 
This reverts commit 06593a072f.

Reason for revert: ROCr query should now be usable in upcoming release.

Change-Id: I2207761ca6af5d585d090bae1af09eb9a8e9bad6


[ROCm/clr commit: a52f5bda8f]
 2022-11-09 10:48:47 -05:00
Julia Jiang a83594277d SWDEV-360748 - Update return value for hipExtGetLinkTypeAndHopCount_Positive_Basic test on Windows 
Change-Id: I0eb984da6e96c2409e1b1bbda61ae694c8bebe94


[ROCm/clr commit: 9d616ad0f4]
 2022-11-07 10:00:32 -05:00
German 50e08f9511 SWDEV-286150 - Switch PAL to 775 version 
Change-Id: I8e64cbe0fea273d4bd6f7d3ebd66feb628fac6b9


[ROCm/clr commit: c1b22840fa]
 2022-11-07 09:54:22 -05:00
Saleel Kudchadker b157c8eeea SWDEV-301667 - Improve logging 
Change-Id: Ifa6da876b85cb503967cf09aac6d477b10db8e63


[ROCm/clr commit: feca11d5e3]
 2022-11-04 18:23:18 -04:00
Saleel Kudchadker 994c9384e7 SWDEV-364052 - Print free memory on queue callback 
Change-Id: I7d15b6f6277326d5f3e9b784a3443f18ec6ba58a


[ROCm/clr commit: 3e465c5ff8]
 2022-11-04 13:19:33 -04:00
Todd tiantuo Li 32b0929ce3 SWDEV-302996 - add gfx1103 Phoenix support - OCL runtime 
Change-Id: Ib6b7efa085b734bda6b5b50c720fe9d93e6dbf95


[ROCm/clr commit: 246429bd68]
 2022-11-04 12:07:22 -04:00
kjayapra-amd 9f4ef7089f SWDEV-329423 - Enable NV3x target ID related changes on mainline. 
Change-Id: I54fafb371976aaa16a452ec65b0c340c5682c264


[ROCm/clr commit: ea0ce45ad3]
 2022-11-03 21:11:24 +00:00
German bc14b745e8 SWDEV-365121 - Use CP DMA for tiny transfers 
Sync between compute and SDMA engines can be very expensive under Windows.
Use CP DMA for tiny transfers (< 1KiB) to avoid syncs and improve performance.

Change-Id: I9db39a2199f7b9e337ed08fd36d9cbc150502f1f


[ROCm/clr commit: 473621c008]
 2022-11-03 10:04:53 -04:00
Tao Sang 441effd3d7 SWDEV-363140 - Support performance counter on gfx11 
Change-Id: I2216f71f4d4db6dd3766333b0c821cb3d35d7851


[ROCm/clr commit: 06573ac92f]
 2022-11-03 07:59:24 -04:00
Laurent Morichetti 6d3939cd49 SWDEV-351980 - Store the correlation ID in the dispatch packet 
The profiler can retrieve this correlation ID to attribute waves to
specific dispatch locations.

Change-Id: I700e8a91219d612f6a2028c0dda0c92753f3526a


[ROCm/clr commit: b043b4f5a2]
 2022-11-02 23:40:10 -04:00
jatang 53f289e0ac SWDEV-293816 - Only dispatch Counter Aql packet when it is created successfully. 
Change-Id: Ic8a7374d19012754c6de5d483ec5d07a56661d1b


[ROCm/clr commit: 166ef6b622]
 2022-10-31 15:11:41 -04:00
Saleel Kudchadker 0d1d52d249 SWDEV-301667 - Use queue base address for AQL log 
Change-Id: Ie525991b341deeb93ab3cb2d77550e8e1d87ee20


[ROCm/clr commit: cf1fd98074]
 2022-10-31 13:54:49 -04:00
German c04e2300c8 SWDEV-363074 - Clean-up sync between SDMA and compute 
HIP can't rely on the resource tracking, used in OCL and requires different explicit sync.
Make sure ROCCLR syncs compute only when SDMA is used and vise versa.
The new logic will allow to enable CPDMA without unnecessary waits.

Change-Id: Ib9d1788cfd5afa5ea2fec4c96a37d8b9c4d0059d


[ROCm/clr commit: ff6b4db70b]
 2022-10-31 10:02:01 -04:00
German c15bee3157 SWDEV-363074 - Enable resource cache for SVM 
Blender creates and destroys big allocations during the benchmark.
That causes big delays, because vidmm has to page-in/page-out memory.

Change-Id: I2baf4545807127406e3d2870a7581ff9ae7bcdb5


[ROCm/clr commit: dc4ad8c99c]
 2022-10-27 16:39:29 -04:00
Ajay 09e359f925 SWDEV-357207 - Linux: enable StreamWrite/Wait APIs in deviceAttribute 
Change-Id: I593df2dbb7a0233dd42f8d873510d85bbb27feff


[ROCm/clr commit: 261620a2ef]
 2022-10-21 12:44:35 -04:00