76333 Commits

Author SHA1 Message Date
Maisam Arif 6a76f8a705 Bump Version to 24.6.5.0
Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>
Change-Id: I93d6d397bd8d647f472017c28101dabe9ff8199b
2024-09-20 02:53:45 -05:00
Maisam Arif cf183c9d10 Bump Version to 24.6.5.0
Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>
Change-Id: I93d6d397bd8d647f472017c28101dabe9ff8199b


[ROCm/amdsmi commit: 6a76f8a705]
2024-09-20 02:53:45 -05:00
gabrpham c9a489d437 Moved partition_id from static --asic-info to static --partition.
partition_id also removed from the `amdsmi_asic_info_t` struct and
supporting API has been added for querying partition information.

Signed-off-by: gabrpham <Gabriel.Pham@amd.com>
Change-Id: Id5a6291a77d11bb97a1c7a200fc465898e86e081
2024-09-20 03:48:42 -04:00
gabrpham 0fd0b46b7f Moved partition_id from static --asic-info to static --partition.
partition_id also removed from the `amdsmi_asic_info_t` struct and
supporting API has been added for querying partition information.

Signed-off-by: gabrpham <Gabriel.Pham@amd.com>
Change-Id: Id5a6291a77d11bb97a1c7a200fc465898e86e081


[ROCm/amdsmi commit: c9a489d437]
2024-09-20 03:48:42 -04:00
Maisam Arif 3b7f661e71 Moved KFD information to separate structure and API
Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>
Change-Id: If6eaea589edc704cf408d6391b5f2154134035e7
2024-09-20 03:48:42 -04:00
Maisam Arif 82096d7f74 Moved KFD information to separate structure and API
Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>
Change-Id: If6eaea589edc704cf408d6391b5f2154134035e7


[ROCm/amdsmi commit: 3b7f661e71]
2024-09-20 03:48:42 -04:00
Maisam Arif 2cfae06560 [SWDEV-482058] Updated Packaging for offline installs
Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>
Change-Id: I8908451c013fc944645b5b5df3104a2ff73e72bd
2024-09-20 00:55:48 -04:00
Maisam Arif f1eae9f051 [SWDEV-482058] Updated Packaging for offline installs
Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>
Change-Id: I8908451c013fc944645b5b5df3104a2ff73e72bd


[ROCm/amdsmi commit: 2cfae06560]
2024-09-20 00:55:48 -04:00
Justin Williams f2f02aa317 [SWDEV-469278] Lowered PyYAML post install script dependency
Signed-off-by: Justin Williams <Justin.Williams@amd.com>
Change-Id: I764c8bf01e6cb6acb0b3fc1db396707099e5ed12
2024-09-20 00:55:48 -04:00
Justin Williams 90c854ad72 [SWDEV-469278] Lowered PyYAML post install script dependency
Signed-off-by: Justin Williams <Justin.Williams@amd.com>
Change-Id: I764c8bf01e6cb6acb0b3fc1db396707099e5ed12


[ROCm/amdsmi commit: f2f02aa317]
2024-09-20 00:55:48 -04:00
Anusha GodavarthySurya f278ac9ea0 SWDEV-485904 - Fix virtual,physical mem obj leaks
Change-Id: Ie0456b5dcfec206ae54a6aabfc2a15a620cac693


[ROCm/clr commit: 870842201d]
2024-09-19 23:04:20 -04:00
Anusha GodavarthySurya 870842201d SWDEV-485904 - Fix virtual,physical mem obj leaks
Change-Id: Ie0456b5dcfec206ae54a6aabfc2a15a620cac693
2024-09-19 23:04:20 -04:00
James Xu be18b69b33 Skip missing vram_str_path and sdma_str_path if sysfs files not created when passing some, but not all, GPUs to a docker image.
- This fix addresses SWDEV-456049 and probably SWDEV-442181 which
	have the same apparent root cause of an early exiting
	loop while enumerating GPU stats

Change-Id: I517329e06fa2c53205d8b6e002895e648ebf521c


[ROCm/rocm_smi_lib commit: 35496cabc4]
2024-09-19 16:53:37 -04:00
James Xu 35496cabc4 Skip missing vram_str_path and sdma_str_path if sysfs files not created when passing some, but not all, GPUs to a docker image.
- This fix addresses SWDEV-456049 and probably SWDEV-442181 which
	have the same apparent root cause of an early exiting
	loop while enumerating GPU stats

Change-Id: I517329e06fa2c53205d8b6e002895e648ebf521c
2024-09-19 16:53:37 -04:00
Shweta Khatri ff6e1b44bf hsakmt: openclose.c: Fix Coverity reported warnings
Add check before close to prevent closing invalid file descriptors

Change-Id: Ie1d50e0d55159512a14a70c1e4be058218aae668
2024-09-19 19:44:53 +00:00
Shweta Khatri 848340186f hsakmt: openclose.c: Fix Coverity reported warnings
Add check before close to prevent closing invalid file descriptors

Change-Id: Ie1d50e0d55159512a14a70c1e4be058218aae668


[ROCm/ROCR-Runtime commit: ff6e1b44bf]
2024-09-19 19:44:53 +00:00
James Xu f3664fd124 rocr: Add nullptr check in IterateExecutables
When an entry is deleted from the array, it's set to nullptr
but not removed. Most other functions that
iterate over the array check if the entry is nullptr
but this loop in IterateExecutables did not.

Change-Id: I763b361eea59f6df201bb86ead0234e95f2cf79c
2024-09-19 19:44:53 +00:00
James Xu 94ee49bdbd rocr: Add nullptr check in IterateExecutables
When an entry is deleted from the array, it's set to nullptr
but not removed. Most other functions that
iterate over the array check if the entry is nullptr
but this loop in IterateExecutables did not.

Change-Id: I763b361eea59f6df201bb86ead0234e95f2cf79c


[ROCm/ROCR-Runtime commit: f3664fd124]
2024-09-19 19:44:53 +00:00
David Yat Sin 7f3dcd4e0b rocr: Add extended fine-grain memory on host
Change-Id: Id9317fee89b51a5097459255e0a3092820eff430
2024-09-19 19:44:53 +00:00
David Yat Sin c6bb1eef90 rocr: Add extended fine-grain memory on host
Change-Id: Id9317fee89b51a5097459255e0a3092820eff430


[ROCm/ROCR-Runtime commit: 7f3dcd4e0b]
2024-09-19 19:44:53 +00:00
David Yat Sin 0af7a54ebe rocr: Return err when freeing invalid pointer
Return false if trying to free a NULL pointer (or invalid size)
internally in ROCr. This is to detect errors within ROCr when trying
to free NULL pointers. If a user of ROCr tries to free a NULL
pointer, this condition should be caught at the beginning of the
Runtime::FreeMemory(...) function and return HSA_STATUS_SUCCESS. This
matches the behavior of the free(...) or delete functions that
silently ignores calls when the passed a NULL pointer.

Change-Id: I84bc26928b35023e19cd9f214b42c6ee9508029c
2024-09-19 19:44:53 +00:00
David Yat Sin 098bccbda4 rocr: Return err when freeing invalid pointer
Return false if trying to free a NULL pointer (or invalid size)
internally in ROCr. This is to detect errors within ROCr when trying
to free NULL pointers. If a user of ROCr tries to free a NULL
pointer, this condition should be caught at the beginning of the
Runtime::FreeMemory(...) function and return HSA_STATUS_SUCCESS. This
matches the behavior of the free(...) or delete functions that
silently ignores calls when the passed a NULL pointer.

Change-Id: I84bc26928b35023e19cd9f214b42c6ee9508029c


[ROCm/ROCR-Runtime commit: 0af7a54ebe]
2024-09-19 19:44:53 +00:00
David Yat Sin 561c44a4a9 rocr: extend agents_allow_access support VMM
Extend hsa_amd_agents_allow_access API to handle memory allocations done
via VMM APIs.

Change-Id: I4ae51d3e42dd104e98d513b1da86133d312a7203
2024-09-19 19:44:53 +00:00
David Yat Sin 5322dd3116 rocr: extend agents_allow_access support VMM
Extend hsa_amd_agents_allow_access API to handle memory allocations done
via VMM APIs.

Change-Id: I4ae51d3e42dd104e98d513b1da86133d312a7203


[ROCm/ROCR-Runtime commit: 561c44a4a9]
2024-09-19 19:44:53 +00:00
David Yat Sin 8f1b05660a rocr: refactor VMemorySetAccess function
Refactor VMemorySetAccess so that it can be re-used in the following
patch.

Change-Id: I341241da7a59724bb3611172f0d26b0689d7bb46
2024-09-19 19:44:53 +00:00
David Yat Sin c34ba41fe4 rocr: refactor VMemorySetAccess function
Refactor VMemorySetAccess so that it can be re-used in the following
patch.

Change-Id: I341241da7a59724bb3611172f0d26b0689d7bb46


[ROCm/ROCR-Runtime commit: 8f1b05660a]
2024-09-19 19:44:53 +00:00
David Yat Sin 588a5a2fd3 rocrtst: Limit pool sizes to 2GB on emulator
Individual simple tests such as CPUAccessToGPUMemoryTest are taking
several hours on emulators as the total amount of VRAM keeps increasing.
Limit the pool sizes to 2GB, only on emulator.

Change-Id: I4b33e8549f89413da255731e6748f606ca64a663
2024-09-19 19:44:53 +00:00
David Yat Sin a0ccfce615 rocrtst: Limit pool sizes to 2GB on emulator
Individual simple tests such as CPUAccessToGPUMemoryTest are taking
several hours on emulators as the total amount of VRAM keeps increasing.
Limit the pool sizes to 2GB, only on emulator.

Change-Id: I4b33e8549f89413da255731e6748f606ca64a663


[ROCm/ROCR-Runtime commit: 588a5a2fd3]
2024-09-19 19:44:53 +00:00
Sam Wu 5506b9af7a Update rocm-docs-core to 1.8.0
Updating past rocm-docs-core 0.x.x requires Python 3.10 (specified in .readthedocs.yaml config)

Change-Id: I3421aa92cda62d48a1466046f1fdeb0a3abf3ef7
2024-09-19 19:44:53 +00:00
Sam Wu 01c1e91c70 Update rocm-docs-core to 1.8.0
Updating past rocm-docs-core 0.x.x requires Python 3.10 (specified in .readthedocs.yaml config)

Change-Id: I3421aa92cda62d48a1466046f1fdeb0a3abf3ef7


[ROCm/ROCR-Runtime commit: 5506b9af7a]
2024-09-19 19:44:53 +00:00
Tony Gutierrez 4bf102dc6b rocr/aie: Support VMEM handle creation
Adds support for AllocateMemoryOnly inside XDNA driver.

Move the IsLocalMemory() check inside the KFD driver
since the XDNA driver can, and needs to, create handles
on system memory buffer objects.

Changed handle variable name from thunk_handle to user_mode_driver_handle,
which is more representative if we support non-GPU drivers.

Change-Id: I95db9d575afd1ab0ff2de74cea5175d9a12a721b
2024-09-19 19:44:53 +00:00
Tony Gutierrez 29de0df770 rocr/aie: Support VMEM handle creation
Adds support for AllocateMemoryOnly inside XDNA driver.

Move the IsLocalMemory() check inside the KFD driver
since the XDNA driver can, and needs to, create handles
on system memory buffer objects.

Changed handle variable name from thunk_handle to user_mode_driver_handle,
which is more representative if we support non-GPU drivers.

Change-Id: I95db9d575afd1ab0ff2de74cea5175d9a12a721b


[ROCm/ROCR-Runtime commit: 4bf102dc6b]
2024-09-19 19:44:53 +00:00
Tony Gutierrez 54a459c05c rocr/aie: Allocate AIE queue's ring buf
Change-Id: I799a8223d695ec5c0ea2eaea012bc1b5d877e103
2024-09-19 19:44:53 +00:00
Tony Gutierrez f58a06c656 rocr/aie: Allocate AIE queue's ring buf
Change-Id: I799a8223d695ec5c0ea2eaea012bc1b5d877e103


[ROCm/ROCR-Runtime commit: 54a459c05c]
2024-09-19 19:44:53 +00:00
Kent Russell 2a9572dda0 hsakmt: Fix BUILD_SHARED_LIBS usage
If BUILD_SHARED_LIBS doesn't get set at all, other projects importing
hsakmt may have an inconsistent state regarding SHARED vs STATIC.
Instead of setting the option and setting it as an option(), just detect
the variable and use it

Change-Id: I9d5a5fc6049ca5351f5e7c63d38ee9bfcb89bdad
Signed-off-by: Kent Russell <kent.russell@amd.com>
2024-09-19 19:44:53 +00:00
Kent Russell 7e130369db hsakmt: Fix BUILD_SHARED_LIBS usage
If BUILD_SHARED_LIBS doesn't get set at all, other projects importing
hsakmt may have an inconsistent state regarding SHARED vs STATIC.
Instead of setting the option and setting it as an option(), just detect
the variable and use it

Change-Id: I9d5a5fc6049ca5351f5e7c63d38ee9bfcb89bdad
Signed-off-by: Kent Russell <kent.russell@amd.com>


[ROCm/ROCR-Runtime commit: 2a9572dda0]
2024-09-19 19:44:53 +00:00
Kent Russell 3b61f75f49 hsakmt: Remove unused functions
The fmm_node_[added|removed] functions were added in the initial FMM
support, but weren't used. Remove them now since no one's referencing
them

Change-Id: I1e46e57294a72012227b38f46c7099de0b9263be
Signed-off-by: Kent Russell <kent.russell@amd.com>
2024-09-19 19:44:53 +00:00
Kent Russell 0dafd65205 hsakmt: Remove unused functions
The fmm_node_[added|removed] functions were added in the initial FMM
support, but weren't used. Remove them now since no one's referencing
them

Change-Id: I1e46e57294a72012227b38f46c7099de0b9263be
Signed-off-by: Kent Russell <kent.russell@amd.com>


[ROCm/ROCR-Runtime commit: 3b61f75f49]
2024-09-19 19:44:53 +00:00
David Yat Sin 0f241d4061 hsakmt: Add debug prints to trace mem allocations
Add extra debug prints to trace memory alloc and register

Change-Id: I03d8d7d415565916a8336db6e7063bb7d4cb9102
2024-09-19 19:44:53 +00:00
David Yat Sin 75e07cc34d hsakmt: Add debug prints to trace mem allocations
Add extra debug prints to trace memory alloc and register

Change-Id: I03d8d7d415565916a8336db6e7063bb7d4cb9102


[ROCm/ROCR-Runtime commit: 0f241d4061]
2024-09-19 19:44:53 +00:00
Ramesh Errabolu 71fa3fa19b Do not check default value of SVM attribute Granularity
Change-Id: I3cf97fc551259c873351cfd22fc83e8615cc3e56
2024-09-19 19:44:53 +00:00
Ramesh Errabolu 72dc577172 Do not check default value of SVM attribute Granularity
Change-Id: I3cf97fc551259c873351cfd22fc83e8615cc3e56


[ROCm/ROCR-Runtime commit: 71fa3fa19b]
2024-09-19 19:44:53 +00:00
Ranjith Ramakrishnan d60f56ab32 Remove license file from hsa-rocr-devel package
License file is already there in hsa-rocr package .Devel package do not need the same

Change-Id: I08cceeb169d0c061078cd495342f78c089087f0d
2024-09-19 19:44:53 +00:00
Ranjith Ramakrishnan 2e97c0d3bd Remove license file from hsa-rocr-devel package
License file is already there in hsa-rocr package .Devel package do not need the same

Change-Id: I08cceeb169d0c061078cd495342f78c089087f0d


[ROCm/ROCR-Runtime commit: d60f56ab32]
2024-09-19 19:44:53 +00:00
Tony Gutierrez a851f73da5 rocr/aie: Init mem regions for AIE agents
Change-Id: If180bdbcb3eb659f0d05a710526864494316d7a9
2024-09-19 19:44:53 +00:00
Tony Gutierrez db03f1130f rocr/aie: Init mem regions for AIE agents
Change-Id: If180bdbcb3eb659f0d05a710526864494316d7a9


[ROCm/ROCR-Runtime commit: a851f73da5]
2024-09-19 19:44:53 +00:00
Tony Gutierrez 6abb993f65 rocr/aie: Add AMD AIE Embedded Runtime vendor packets
Adds support for the packet interface for interacting with
the Embedded Runtime (ERT) on AIE agents. The ERT is what
interprets command packets send to the AIE agent work
queues.

Change-Id: Id28fb98056b2c046354c446bdc9568d74385bea1
2024-09-19 19:44:53 +00:00
Tony Gutierrez f096174cfc rocr/aie: Add AMD AIE Embedded Runtime vendor packets
Adds support for the packet interface for interacting with
the Embedded Runtime (ERT) on AIE agents. The ERT is what
interprets command packets send to the AIE agent work
queues.

Change-Id: Id28fb98056b2c046354c446bdc9568d74385bea1


[ROCm/ROCR-Runtime commit: 6abb993f65]
2024-09-19 19:44:53 +00:00
Tony Gutierrez 931733d51a rocr/aie: Add support for creating AIE queue context
Adds support for initialzing the XDNA driver so that
a hardware context can be created for an AIE queue.

Right now this initializes the device heap in the driver,
gets the relevant tile parameters for the AIE agent,
and creates a hardware context that backs the AIE queue.

Change-Id: Ib90e1bc67a8637f6db3ff2bebe34677843796417
2024-09-19 19:44:53 +00:00
Tony Gutierrez 1103441d22 rocr/aie: Add support for creating AIE queue context
Adds support for initialzing the XDNA driver so that
a hardware context can be created for an AIE queue.

Right now this initializes the device heap in the driver,
gets the relevant tile parameters for the AIE agent,
and creates a hardware context that backs the AIE queue.

Change-Id: Ib90e1bc67a8637f6db3ff2bebe34677843796417


[ROCm/ROCR-Runtime commit: 931733d51a]
2024-09-19 19:44:53 +00:00