Commit graph

2484 Commits

Autor SHA1 Nachricht Datum
Lancelot SIX 855015377c Add GFX12 trap handler
Given the differences between previous architectures and gfx12, this
patch implements the gfx12 2nd level trap handler in a separate source
file, and adjusts the build system.

Change-Id: I65192ffbbcd66a4f78d2d0c3fb1739a92cac95d4
Signed-off-by: Lancelot SIX <lancelot.six@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-25 12:27:09 -05:00
Sreekant Somasekharan 24463635f9 Initial GFX1201 changes.
Add target gfx1201 to several files.

Change-Id: I5cae7dba00ed58f8fbfa6e7147275bd7d5feaed0
Signed-off-by: Sreekant Somasekharan <sreekant.somasekharan@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-25 12:27:09 -05:00
David Belanger 40cc6559f1 Add Blit shaders for GFX12
For GFX12, the workgroup id is passed in ttmp9 (trap temp register) instead of the scalar register.
Normal shader code (i.e. not priv, not trap handler) can only read the ttmp registers.

Signed-off-by: David Belanger <david.belanger@amd.com>
Change-Id: I42404d8c8c0ee9c746e23879fd30b2d16cfa1787
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-25 12:27:09 -05:00
Shweta.Khatri 4e9647704d Fix soft hang on AQLQueue destruction with a timeout
Add timeout to AQLQueue destructor signal wait to prevent indefinite hang

Change-Id: I6c6c98a7bdd27d39569af1d667aa9aa7e9596535
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-25 12:27:09 -05:00
David Yat Sin 2f05c2a273 Revert "Use pthread_setaffinity_np"
This reverts commit 1df7a44112e45b7fb447926778490f741601219a.

Change-Id: Ib386c8f944b6da0ef68ddd2be3f26013cd36ef5b
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-25 12:27:09 -05:00
David Yat Sin 1cee8656df Revert "Use pthread_attr_setaffinity_np when available"
This reverts commit ef95ccf81e59b8608861e8f2f256d981eee19df7.

Reason for revert: Causing performance regressions on some systems

Change-Id: I82951350cafbd57c495852d6f90023a3373f04f6
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-25 12:27:09 -05:00
David Belanger 2f14acd9c1 Initial GFX12 changes.
Add target gfx1200 to several files.
Add cases for GFX12 in a few switch statements.

Signed-off-by: David Belanger <david.belanger@amd.com>
Change-Id: Ib90032f5b9d5a3306060f13a43d970108a1399df
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-25 12:27:09 -05:00
Ranjith Ramakrishnan 696d8fae9e Static package generation for hsa
Generate static package by combining binary and dev components.
Binary and dev component dependencies are added to the static package dependencies
No dependency to rocprofiler-register
Package name will have suffix static-dev/devel

Change-Id: I2f9680f13dbffc9eb7ced9fa9b28e360c47ebcca
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-25 12:27:09 -05:00
Tony Gutierrez 69ba32fa95 driver: Add a core driver interface component
Add a new driver interface as a core ROCr component.

The driver component provides an interface for ROCr to interact with
agent kernel-model drivers in a generic way. This interface will be used
to interact with the XDNA NPU driver. Eventually, the ROCt library's
functionality should be implemented behind this interface.

For now the interface provides basic queue and memory allocation
for supporting HSA queues and signals and matches the thunk API
closely.

Change-Id: I37ac9f2dcbadc86ce45999f76b0e9ce753fd0c06
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-25 12:16:40 -05:00
Lang Yu 2f50b35daa Simplify APU query
Query APU from thunk instead of parsing device id.

Change-Id: I95efa9e2a94fb979eaa88042991ee6921abbed7f
Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-25 12:16:40 -05:00
Chris Freehill 662f6817d7 Changes to build ROCr & thunk (optionally tests) in rocr-runtime repo
Create a new top-level CMakeLists.txt file to control building thunk
and ROCr. kfdtest and rocrtest are built separately.

Most of the cmake code that existed for thunk, ROCr, rocrtst and kfdtest
still reside in their respective CMakeLists.txt files, except the
CPack packaging directives which have been moved to the top-level
CMakeLists.txt.

Change-Id: I1a537359029504af8b1abb324bc6f0d75d98471e
2024-06-24 14:26:21 -05:00
David Belanger 8917561625 Update AtomicIncIsa shader for GFX12
Minor instructions changes for GFX12.

Signed-off-by: David Belanger <david.belanger@amd.com>
Change-Id: Iab2c430bb5d7d8fa2b166d07fd33ea15aca3a5cd
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
James Zhu 338721c24a PC Sampling: Temporarily check KFD_IOCTL_MINOR_VERSION 16
Since PC Sampling is still under experiment, we can't
bump KFD_IOCTL_MINOR_VERSION to enable pc sampling.
KFD_IOCTL_MINOR_VERSION 16 already includes all pc sampling
code, so use version 16 to enable pc sampling implicitly for
customer to try-out this new feature.
Need update the version accordingly when pc sampling upstream.

Change-Id: I65840128f94e8f347c0617971c0aa4b7e478691a
Signed-off-by: James Zhu <James.Zhu@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
David Belanger 9665499f83 Update ScratchCopyDwordIsa shader for GFX12
Minor instructions changes for GFX12.

Signed-off-by: David Belanger <david.belanger@amd.com>
Change-Id: I57cca6393d4b4aae869a2bc9862d75eef1f29ed7
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
David Belanger bc3f39a39e kfdtest: Update ReadMemoryIsa shader for GFX12
Minor instructions changes.

Change-Id: Iaa12763c6f7835aa658dbfb121e4963424b16745
Signed-off-by: David Belanger <david.belanger@amd.com>
Signed-off-by: Sreekant Somasekharan <sreekant.somasekharan@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
David Belanger 39f4fbee10 Update CopyOnSignal shader for GFX12
Minor instructions changes for GFX12.

Change-Id: I78a37fa37950b378cdd2a1618c71c97c6ba66aac
Signed-off-by: David Belanger <david.belanger@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Philip Yang 6e6f445f75 libhsakmt: Update contiguous memory support ioctl version
KFD ioctl version is 1.16 on upstream for contiguous memory support.

Remove pc_sampling version, should be added after pc_sample upstream.

Change-Id: I6e6c3340bc8e371d68dd7741b02578be2fdef801
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Philip Yang e076a4ee82 kfdtest: Add RDMATest ContiguousVRAMAllocation
Update amdp2ptest.h to sync with the same file from rdma test driver
folder.

Add ContiguousVRAMAllocation to verify rdma get pages will get
contiguous VRAM pages, skipped RDMA getpages if amdp2ptest.ko is not
loaded.

Change rdma buffer mmap with MAP_SHARED flag, because MAP_PRIVATE goes
to COW path, which requires mmap the entire vma and cannot support
multiple sg nents.

Change-Id: I5fbb1902251f1454616d4404a4b048a88996d4f7
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Philip Yang 9d9fbceefb rdmatest: Fix amdp2ptest driver mmap
mmap system call parameter vma->vm_start, vm_end is mmap virtual address
start, end, vma->vm_pgoff is rdma buffer GPU address, which is used to
find the sg_table dma_address.

Handle multiple sg table nents case because sg->length is limited to max
2GB.

Change-Id: I677dd6662ee58f0b5c93f8eef32b7009e1e890d8
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Philip Yang c98a8dc179 libhsakmt: Add missing CHECK_KFD_OPEN in APIs
The application may use parent process KFD handle or invalid KFD handle,
add CHECK_KFD_OPEN in all APIs to catch this application bug earlier
without calling to KFD.

Change-Id: I0391e91eeca8e6752fc9c23f0742445b823ea9b0
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
David Yat Sin a31e84eaef libhsakmt: Add alignment for memory allocations
New API to support optional alignment parameter for memory allocations.
The alignment should be larger than or equal to page size and a power
of 2.

Change-Id: Ic3fec43b3c4281f74dd33a57ab4143dcf76e1186
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Jesse Zhang 1abd02af32 kfdtest: fix MigrateLargeBufTest issue
Since the amdgpu driver commit 1f4ac94b59aebebf.
https://lore.kernel.org/all/a121a72c-b441-4f42-94a3-4597b7f19e7d@amd.com/T/
gtt and vram are available for compute.
So, the vramSize obtained by function GetSysMemSize is actually about 50% system memory.
But small APUs don't have large system memory, and kernel memory limit is smaller for them.
Therefore, it will fail to register SVM Range for SysBuffer and SysBuffer2.

Example:
  System Memory size: 3373M   Kernel memory limit:1791M
  VRAM Memory Size: 256M    GTT Memory Size: 1686M

Signed-off-by: Jesse Zhang <Jesse.Zhang@amd.com>
Change-Id: Ib3826933100ab7b432cb476caaf2d91cc9cdb948
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Yiyang Wu 4063c7a285 kfdtest: Allow linking libLLVM dynamic library rather than separate components
Change-Id: Idc531ab35924f856600049f7f0101d1141ebcd51
Signed-off-by: Yiyang Wu <xgreenlandforwyy@gmail.com>
Signed-off-by: Kent Russell <kent.russell@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Lang Yu 4844a70d94 libhsakmt: Prevent hsaKmtRegisterMemory* from registering non-userptr
hsaKmtRegisterMemory* can only register OS allocated userptr.

v2: Apply changes to all hsaKmtRegisterMemory* stuff.(Philip)

v3: Unlock aperture->fmm_mutex to aviod deadlock.

Suggested-by: Felix Kuehling <Felix.Kuehling@amd.com>
Change-Id: I1045af7edb4da8206cb878f64c0176ba4fc59f60
Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Your Name 3f00c88910 rdmatest: Add rdma contiguous memory allocation test
Update CMakeLists.txt to use Thunk pkgconfig.

Add rdma contiguous memory allocation test, to verify if KFD rdma get
pages to pin buffer on contiguous VRAM pages.

Change-Id: I7cc617fc083ce1998c214c327c130f033ce41d6f
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Your Name 88dabfcc04 rdmatest: Update amdp2ptest driver
Update Makefile for newer kernel version, and support build with dkms
amdgpu driver. Use symbol_request to get KFD peerdirect interface.

Sync up with KFD peerdirect interface changes, remove the free callback
which is not used any more.

Change-Id: I01d8906d9ffa427a058a26e88e36f6b80e9e22c2
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Lang Yu a7a712fb36 libhsakmt: Fix improper usage of hsaKmtRegisterMemoryToNodes
It's unnecessary to register non-userptr.

Change-Id: Iefd329578365e036e2fe7e4d5c9c0c3d0976f67c
Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Lang Yu 5eb2a2660c kfdtest: Skip KFDEvictTest, KDFSVMEvictTest, HMMProfilingEvent on small APU
Small APUs use same memory allocation approach with APP APU now,
skip these tests as well.

Change-Id: I13c953cc53da071f6f36af0d4a0153a48ea066fe
Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Lang Yu ae3ede062f libhsakmt: add Integrated property
To differentiate discrete and integrated GPU more flexibly in runtime,
this will aid in querying HSA_AMD_MEMORY_PROPERTY_AGENT_IS_APU
and hipDeviceAttributeIntegrated.

Change-Id: Ic8a6c9aea3b4bd19c4d5f6729af7e64c328fc61d
Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Yiyang Wu 9316b6e4e4 kfdtest: hsaKmtCheckRuntimeDebugSupport should be visible
Change-Id: I03a379ede1c990bd275a4d2a8cb379f228381d03
Signed-off-by: Yiyang Wu <xgreenlandforwyy@gmail.com>
Signed-off-by: Kent Russell <kent.russell@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
David Belanger 24578e10c1 kfdtest: Add exclude list for GFX12
Add test cases excluded from GFX11 to GFX12 list if they are also not
stable on GFX12.

Signed-off-by: David Belanger <david.belanger@amd.com>
Change-Id: Ifeab24f8ea94085250ea86128a3e401479bdb53d
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
David Belanger 8854555ed7 kfdtest: Update PollMemoryIsa shader for GFX12
Minor instructions changes.

Signed-off-by: David Belanger <david.belanger@amd.com>
Change-Id: I40d6aaffd78cf27f7c3b436cea5403d39b5b88ec
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
David Belanger bde8e7a212 Update shader for GFX12
Minor changes to instructions for GFX12.

Signed-off-by: David Belanger <david.belanger@amd.com>
Change-Id: Iac5be900e3755099d83010fb1a2066b4dbb52dda
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
David Belanger e086c383fe kfdtest: Updated KFDCWSRTest for GFX12
Updated ShaderStore shader (used by CWSR test) for GFX12.
Workgroup ID now pass in a different register.
Minor changes for new scope syntax.

Signed-off-by: David Belanger <david.belanger@amd.com>
Change-Id: I6fdabc8b62cba201d7777a736d3d43cfae28ca4c
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Jonathan Kim 3b842c39f1 kfdtest: fix address watch test for GFX12
New watchpoint exception status bits have to been assign to the first 4 least
significant bits so change test verification mask to check against the
first watch point ID accordingly.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: If83950207ea9f66cd230c23e7386a97b3893c2eb
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Jonathan Kim a2e9226784 kfdtest: fixup test traphandler for gfx12
Fix traphandler for KFD debugger testing.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>

Change-Id: Ib8f5aac3d1b99e4463ac56b5f6d5dee2c367c447
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
David Belanger 259a724e21 libhsakmt: Fix VGPR size for GFX12/GFX12.1
Set max size needed for VGPR when doing a CWSR for GFX12 and GFX12.1.

Signed-off-by: David Belanger <david.belanger@amd.com>
Change-Id: Iddefc62f1ad419c6f5ab6a872048457a1dc24037
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
David Belanger 197a6c2e6c kfdtest: Added gfx1201 filter
Initial template for GFX12.0.1.

Change-Id: I5d2be1f594bf057c04f6feee75a80c61a9d7e4a8
Signed-off-by: David Belanger <david.belanger@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
David Belanger 7d2c78a37d kfdtest: Add support for GFX12
Added FAMILY_GFX12 code.

Signed-off-by: David Belanger <david.belanger@amd.com>
Change-Id: I26f01055b3c8732b4b6e1195d34533d9f89032d2
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
David Belanger 09744e4959 kfdtest: Added gfx1200 filter.
Initial template for GFX12.

Signed-off-by: David Belanger <david.belanger@amd.com>
Change-Id: I552374bfcc0dd6272d170df85d36d0dbca0196d5
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
James Zhu 5786dbbb76 libhsakmt: update KFD ioctl minor version
Since PC Sampling not upstream yet, so use 1.16 for
contiguous VRAM allocation, and 1,17 for pc sampling.

Change-Id: Ib5d22e8f386ce7fe3f7111485b9632b61227e539
Signed-off-by: James Zhu <James.Zhu@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
James Zhu 1087dea925 kfdtest: skip test when PC Sampling is not supported by ASIC
Skip test when PC Sampling is not supported by ASIC.

Change-Id: I6f9be0bdaed66e51052723b6df6908079470cefb
Signed-off-by: James Zhu <James.Zhu@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Jonathan Kim 206db80a56 libhsakmt: fix pc sampling return of functions
C Error returns are positive in user space and should check against errno
instead.
Fix declaration of return to type HSAKMT_STATUS.
KFD IOCTL should handle size return when querying capabilities so return
size to caller unconditionally.
Clean up error translations per function so that it's stylistically
clear.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: Ic37390425f370c7ad88f9ed014444decf19383a3
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Kent Russell 371d078226 kfdtest.exclude: Fix blacklist
We need : to end each subtest, except for the last entry.

Change-Id: I9515d90703c9679e06a4acd124883540c1d5b832
Signed-off-by: Kent Russell <kent.russell@amd.com>
Signed-off-by: Chris Freehill <cfreehil@amd.com>
2024-06-24 14:26:21 -05:00
Chris Freehill 79e4eda0b6 Merge 'thunk/integrate-into-rocr' into integrate-libhsakmt 2024-05-02 21:52:49 -05:00
David Yat Sin d2d95a8948 rocrtst: add test for contiguous mem allocations
This test may fail when run on non-upstream versions of KFD as this
feature will not be upstreamed.

Change-Id: I7131e1f50984739c0df12e4c9afe790bd7e4cdfa
2024-04-30 17:42:15 -04:00
David Yat Sin ac5fb8be9e Temporary: Do not early release mutex when not ganging
It seesm the Release() function is not reliable and can cause segfaults.
This is a temporary work-around until the Release() function is fixed.

Change-Id: I95470a800c6153673e4b8f4fe46a646903325074
2024-04-30 17:07:39 -04:00
Chris Freehill 11fd5c2562 Prepare for integration into rocr
Change-Id: I6102b9910dbb9d09e09bb262a03c5c0ad4ce66f4
2024-04-30 09:01:09 -05:00
David Yat Sin 57b93e02a4 Use pthread_attr_setaffinity_np when available
If pthread_attr_setaffinity_np function exists use it instead of
pthread_setaffinity_np as pthread_setaffinity_np seems to fail to set
the affinity settings on some systems.

Change-Id: Icd8b17039699ac10d9cd5c4dbb6ac44630673949
2024-04-29 15:02:54 +00:00
David Yat Sin b6829f7a72 Bump HSA_AMD_INTERFACE_VERSION_MINOR
Bumping HSA_AMD_INTERFACE_VERSION_MINOR version to 5 to account for
previously added GPU agent query: HSA_AMD_AGENT_INFO_MEMORY_PROPERTIES

Change-Id: Ic8cfdcfb7bad6f3d1e0b3d68f505a62074fc26b9
2024-04-29 12:55:18 +00:00