821ae6a103
SWDEV-426272 - hipPointerGetAttribute shouldn't segfault when host ptr is passed
...
Check the pointer if its present in the arrayset before trying to dereference
it as it can cause access violation if the pointer is allocated using malloc
Change-Id: Ida72b9015dc22269fc1fbe0728e66e3de29fda3d
2024-02-21 16:37:21 +00:00
1f6d416684
SWDEV-446399 - Fixed segfault in hipMemSetAccess
...
Change-Id: Ia1200d9bee03e8abade211287505f081e635ceec
2024-02-20 18:51:05 -05:00
f138e0d113
SWDEV-443760 - Enable device kern args
...
- Implement workaround to ensure HDP writes are done by writing and
reading the HDP MMIO register.
- Implement the same workaround for graphs, we no longer need sentinel
write/readback
Change-Id: I0d3027b46a1f61131ec62e3c8c669ff5184fa6b2
2024-02-20 02:03:14 -05:00
81b8598af9
SWDEV-301667 - Cleanup code and better log
...
Change-Id: Ie2345264e84026156a9f81b421eed3cf4aeeeffc
2024-02-19 05:42:47 -05:00
7d09e1abed
SWDEV-444767 - Fix graph tests for context change between Inst & launch with DEBUG_CLR_GRAPH_PACKET_CAPTURE
...
When graph is Instantiate on device 0 graph and launch on device1 switch to command creation and enqueue during launch.
Change-Id: Ied34dc99b2a776130d1354ed3830c6ccab9068e4
2024-02-14 17:02:36 +00:00
d1b75659f2
SWDEV-445539, SWDEV-445549 - Restore kernel launch prof data.
...
Issue is due to INIT_hipLaunchKernel_CB_ARGS_DATA turns out to be empty when we generate new hip_prof_str.h with since
extra space introduced by https://gerrit-git.amd.com/c/compute/ec/clr/+/846275/77/hipamd/src/hip_module.cpp#667
Change-Id: Id251956c1663483cc7ab11a3722769cb5b0d3664
2024-02-14 00:05:41 -05:00
243555959f
SWDEV-438299 - Remove unused variables
...
Change-Id: Id2908fe851d566ec2515161090f1e21b60cf6447
2024-02-13 19:39:11 -05:00
4f6b0da028
SWDEV-311271 - Fix Windows mempool management
...
Windows path still uses multi threading implementation. Hence, in graphs
all nodes are executed in a queue thread and that requires to manage
mempool in the queue thread. However, the spec allows to destory memory,
allocated in a graph, outside of the graph's execution. That may cause
mempool management to go out of sync.
Change-Id: I0ffb2244b3cb720455ed44d1b3e2487fa8959a77
2024-02-13 17:16:06 -05:00
853abeb75e
SWDEV-445013 - During CaptureAQLPackets correct sentinal value to copy integer size bytes
...
Read and write int bytes sentinal value to dev_ptr or PCIE connected devices at the tail end of the kernarg surface.
Change-Id: I993d552ac872b3cd56aef4746c4d1d92c58d38b4
2024-02-13 07:05:57 +00:00
4d61237a6f
SWDEV-445576 - Return error when hipMemPoolDestroy is invoked on the default mempool
...
Change-Id: I7a7ab499f781cf7e92cad1ae39a40b4f37f62653
2024-02-12 12:48:21 +00:00
5e6ad190ff
SWDEV-445221 - Fix the memory reported for hipMemPoolAttrReservedMemHigh
...
Change-Id: I894a836643612c73feb322f29a9236ada4d8cbda
2024-02-08 23:40:51 +00:00
0479cdb3dd
SWDEV-437702 - implement hipGetProcAddress
...
This should be used in place of dlsym or GetProcAddress (linux and windows respectively)
Change-Id: I5501b538e03892e8e5a2282678d848fcaf21d911
2024-02-08 11:26:02 -05:00
d6bc40e822
SWDEV-445084 - Add DEBUG_CLR_GRAPH_PACKET_CAPTURE support for hipGraphInstantiateWithFlags/Params
...
Change-Id: I5096b4c8d73d1faf972dfd23ab86a53d888946c4
2024-02-08 04:55:53 -05:00
ca0b50c9ca
SWDEV-444558 - SWDEV-444418 - Fix capturing of AQL packets when kernel arg size is 0
...
When graph doesn't have kernel nodes.
Change-Id: I6b3b476654d7eedc9ff0cec4b7269168aa115360
2024-02-08 06:12:16 +00:00
f964975db0
SWDEV-421025 - Graph Instantiate with Params API Update
...
Change-Id: I3ed821ced02420858d360e8dab5e1e931c350c7e
2024-02-07 11:35:21 -05:00
7d5b4a8f7a
SWDEV-437832 - Changes to update host unified memory and iommuv2 flags.
...
Change-Id: I88998cf57c21fc446fa28e250f826c607923670b
2024-02-07 06:27:47 -05:00
a6841e3f43
SWDEV-400448 - SWDEV-392872 - Optimize to add lock only if its null stream
...
Change-Id: Ia905109a295614e8ce4e2bd24261c86ba8138178
2024-02-07 04:51:54 -05:00
a1b2cbe44e
SWDEV-439637 - Updated to compile with clang compiler
...
Change-Id: Ib0a8e1cc007f083fb1d1f4363cf89ba76ad3c4f2
2024-02-06 23:57:13 -05:00
5303cd2740
SWDEV-439637 - Fix undefined symbol error hipTexRefGetArray & hipTexRefGetBorderColor (github id- #3382 )
...
Change-Id: Iae1956779e84759f23011151b60a01cef0d0ef32
2024-02-06 03:04:07 -05:00
8d265838cb
SWDEV-400448 - SWDEV-392872 - Add lock to NullStream
...
Fetching null stream's logic has changed earlier from amd::HostQueue
to hip::Stream. This seem to cause some timing difference between
checking for null stream and creating it due to which issues are
observed in multithreaded applications using default stream.
Change-Id: Ie02365dec537275d23a1d225de9811e2fd3a9c55
2024-02-05 02:58:23 -05:00
1a3901fa49
SWDEV-421025 - hipGraphInstantiateWithParams API changes
...
Change-Id: Ib07d4dd1698220b68ed27f91d58d3bd315a8804c
2024-02-05 05:08:11 +00:00
ceffb24416
SWDEV-421025 - Hip Launch Attribute ID/Value datatype for rocprof
...
Change-Id: Icfcd1ecf2f2956ba015820ba250f06b438169a6a
2024-02-05 05:08:11 +00:00
ae0368d12d
SWDEV-422207 - Enable DEBUG_CLR_GRAPH_PACKET_CAPTURE environiment variable
...
Change-Id: I9bf72b9c1a56980352109bd4d42b54ecb2d1b8f9
2024-02-05 05:08:11 +00:00
5b72582474
SWDEV-443018 - Restrict input range for value argument in hipPointerSetAttribute
...
Change-Id: I17b6448bf49aed49315a0b4c1426dcbe011735d9
2024-02-05 05:08:11 +00:00
a2a8dad980
SWDEV-442719 - Set CMP0074 to NEW to make sure find_package works correctly
...
If a system has LLVM installed, `find_package` could choose that one
even if we set `HIP_LLVM_ROOT`. `LLVM_ROOT` is ignored because of this
CMake policy is set to `OLD` by default.
Change-Id: I18fa0453afe170c229e92d6ddc386b43eb0c44f6
2024-02-05 05:08:11 +00:00
60549b101c
SWDEV-438503 - Always report hipDeviceAttributeUnifiedAddressing as true.
...
Change-Id: I3119bd8f99c079280bde1ba32b3d26b474899e7d
2024-02-05 05:08:11 +00:00
f28b1f1f1d
SWDEV-443109 - Fix for CXX compiler check in hip-config-amd.cmake
...
Change-Id: Ib6b8bc59a889ad2e4f440e847d9f2e7e8fb2cda2
2024-02-05 05:08:11 +00:00
d80168eb87
SWDEV-295298: hide warp sync builtins with a macro in ROCm 6.1
...
Change-Id: Ie2efd233c0bcf8ad0e06223ee854fe4bd1060443
2024-02-05 05:08:11 +00:00
9fd8b4d3f1
SWDEV-440029 - Fix hipMemcpyFrom/ToSymbol directions allowed
...
Change-Id: Ib0b3d737cd5d297ac2d3c1ba549902b96e29c18c
2024-02-05 05:08:11 +00:00
b53fcb4d02
SWDEV-440718 - Suffix hip bin name with _6.
...
Link to amd_comgr_2.dll
Change-Id: I5d21af0cc91f2082099273f75ebe1c3279b6365d
2024-02-05 05:08:11 +00:00
e9957151f3
SWDEV-439628 - hipGraphExecKernelNodeSetParams to update graph kernel node params with graph performance optimizations.
...
During hipGraphExecKernelNodeSetParams kernel function can also be updated.
Hence size required for kernel parameters differs from what is allocated during graphInstantiation.
So, create new 128KB kernel pool and allocate kernel args from the pool.
If the pool is full create new 128KB pool. Release kernel pools when graph exec object is destroyed.
Change-Id: I9567946d63400c79cbfd4c5439c654c92557ceae
2024-02-05 05:08:11 +00:00
242132caa7
SWDEV-439637 - Fix undefined symbol error(github id- #3382 )
...
Change-Id: Ic7812e21618f48c737ce2c51b22a153099cc5d66
2024-02-05 05:08:11 +00:00
aacdb00759
SWDEV-425605 - remove exe_options, it was unused
...
Change-Id: I0651f6a836d2f063caf651520de96d7675428771
2024-02-05 05:08:11 +00:00
2dac58e372
SWDEV-441482 - Add missing entry point
...
hipDestroyExternalSemaphore_fn was missing initialization
Change-Id: Ieab019afd0a0c62a2aa534fb262436edda3499db
2024-02-05 05:08:11 +00:00
f017777a97
SWDEV-425605 - Add new comgr compile to reloc
...
use AMD_COMGR_ACTION_COMPILE_SOURCE_TO_RELOCATABLE action
to compile source to realoc. Currently we have source->bc,
link->bc and bc->realoc. This new action replaces the
three steps with one.
Change-Id: I8089cbef681e079702fefc2d2085a23bc3578d02
2024-02-05 05:08:11 +00:00
140a075b6a
SWDEV-311271 - Move mempool clean-up into hipEventSynchronize
...
Generic event synchronize can be called from mempool implementation directly.
Change-Id: Id66ab9cdab486390df4dd405a4ac291eb0bb568f
2024-02-05 05:08:11 +00:00
f9545675cb
SWDEV-441937 - Use the disable linemarker option for generating the precompiled header file hipRTC
...
The precompiled header files have hard coded paths in comments. Using the disable linemarker option(-P) will skip the generation of comments
Change-Id: Ifb134052996c343f5405e954784b4b2c286c36b1
2024-02-05 05:08:11 +00:00
cd59d971f0
SWDEV-311271 - Add extra logic to reduce memory usage
...
Add logic to prevent significant grow of reserved memory in mempool.
Change-Id: If034433c26e904e3037200a593ef338896d7f16d
2024-02-05 05:08:11 +00:00
cec5f0b704
SWDEV-420140 - Prefer forward slashes for path seperators in compiler RT path with windows
...
Certain build systems such as NMake seem to interpret backslashes as escapes.
Change-Id: Ifc04d9427148e83084efb1af94a1ce85cb2a6c4e
2024-02-05 05:08:11 +00:00
853b293983
SWDEV-433312 - Return invalidArgument in cuMemSetAccess for hipMemAccessFlagsProtRead
...
Change-Id: I8e5c460ac310ad32849359720a40fc5131d4f32c
2024-02-05 05:08:11 +00:00
baf61b6a02
Revert "SWDEV-425605 - Add new comgr compile to reloc"
...
This reverts commit 551cdcd4cfc1a9bb8b8ebfd315320629baf05655.
Reason for revert: breaks windows-psdb
Change-Id: I8b2fcaf0d0be09a9364e386127cabafb68a67e0d
2024-02-05 05:08:11 +00:00
c4b9442fc4
SWDEV-442126 - Fix use_after_free case in ExtractFatBinaryUsingCOMGR
...
Change-Id: Ie07cf82025b65e4f95d4ce3cf46a6b8081f29b6a
2024-02-05 05:08:11 +00:00
b366a7c992
SWDEV-437832 - Adding device property to check if the device is accelerator.
...
Change-Id: I8349e99c03422c268bbb60a8c143bd492d9cec09
2024-02-05 05:08:11 +00:00
0a055f874b
SWDEV-422207 - Added debug env to dump graph during Instantiation
...
Change-Id: Ibde2ae5b8d240f3986bcd168facc513a319c0f17
2024-02-05 05:08:11 +00:00
ea1487d84d
SWDEV-425605 - Add new comgr compile to reloc
...
use AMD_COMGR_ACTION_COMPILE_SOURCE_TO_RELOCATABLE action
to compile source to realoc. Currently we have source->bc,
link->bc and bc->realoc. This new action replaces the
three steps with one.
Change-Id: I6ba551b8d04c7e06f41c4324026e4dcd2db1970f
2024-02-05 05:08:11 +00:00
6af1c419ce
SWDEV-441065 - Fix device accessable host memory copy path
...
Change-Id: Ia6b9a428da7b54819fb3be928a33cdffcd04ecdf
Signed-off-by: Lang Yu <Lang.Yu@amd.com >
2024-02-05 05:08:11 +00:00
a01257f000
SWDEV-431367 - fix float compare for atomicMax/Min where -0.0 < +0.0
...
Change-Id: Ief99aac3775b0a0c6372215e5a0a9ecbbdca4ca5
2024-02-05 05:08:11 +00:00
0f9f8b52da
SWDEV-435750 - Make threadIdx, blockIdx compatible with CUDA
...
Change-Id: If0bd2c2c72e8418e30b5a202d9c10da334eff75e
2024-01-18 02:16:15 -05:00
f40c380cdb
SWDEV-421025 - Graph Kernel Node priority Attribute Set/Get
...
Change-Id: I5c422728aa694c8dabb5cf9bade441101512a249
2024-01-17 12:44:35 -05:00
fb2c100d8c
SWDEV-398109 - Enable smid support for CU mode for gfx10+.
...
Change-Id: Ief1e7c952c3de7e4a547bbdcc9817f51aa4d5d36
2024-01-16 20:13:23 -05:00
Satyanvesh Dittakavi
Ioannis Assiouras
Saleel Kudchadker
Anusha GodavarthySurya
Jaydeep Patel
Ilya Panfilov
German Andryeyev
jiabaxie
Rahul Manocha
kjayapra-amd
Shilei Tian
Todd tiantuo Li
Sameer Sahasrabuddhe
Ajay
Jatin Chaudhary
Ranjith Ramakrishnan
Vikram
Lang Yu