76333 کامیت‌ها
مولف SHA1 پیام تاریخ
Chris Freehill 7effb405f0 Change Debian Architecture from amd64 to any
rocm_smi_lib is not currently known to only compile
on specific architectures.

Change-Id: I209e8baa063e99ebe5ff09eaf0dc6541770aa829
2021-02-01 13:48:38 -06:00
Chris Freehill 9d2e2ffffd Change Debian Architecture from amd64 to any
rocm_smi_lib is not currently known to only compile
on specific architectures.

Change-Id: I209e8baa063e99ebe5ff09eaf0dc6541770aa829


[ROCm/amdsmi commit: 7effb405f0]
2021-02-01 13:48:38 -06:00
gilbertlee-amd 60c74f63fa [TransferBench] Restore some previous fixes - memory leak, PCIe address (#314)
[ROCm/rccl commit: 62e0447e9a]
2021-02-01 09:48:09 -07:00
gilbertlee-amd 62e0447e9a [TransferBench] Restore some previous fixes - memory leak, PCIe address (#314) 2021-02-01 09:48:09 -07:00
Joseph Greathouse 62fee66ff2 SWDEV-270773 - Fix workgroup calculation logic for GWS initialization
The existing workgroup calculation logic for GWS initialization is
incorrect. It tries to add together workgroups across dimensions,
leading to major under-count in 2D and 3D kernels. An (x,y,z) kernel
uses x * y * z blocks, not x + y + z.

In addition, the previous logic was incorrect for the case of launching
a single-threaded kernel. It calculated 0 workgroups, leading to
initializing GWS to -1.

Change-Id: I1bb20a0d5b6e0cc10ac55901c28d8f93aac61c09


[ROCm/clr commit: 54d1d69c0a]
2021-01-31 01:16:48 -06:00
Joseph Greathouse 54d1d69c0a SWDEV-270773 - Fix workgroup calculation logic for GWS initialization
The existing workgroup calculation logic for GWS initialization is
incorrect. It tries to add together workgroups across dimensions,
leading to major under-count in 2D and 3D kernels. An (x,y,z) kernel
uses x * y * z blocks, not x + y + z.

In addition, the previous logic was incorrect for the case of launching
a single-threaded kernel. It calculated 0 workgroups, leading to
initializing GWS to -1.

Change-Id: I1bb20a0d5b6e0cc10ac55901c28d8f93aac61c09
2021-01-31 01:16:48 -06:00
Chris Freehill a9d0e037b5 Add to/correct handling of RDC_EVNT_XGMI_*_THRPUT events
RDC_EVNT_XGMI_[2-5]_THRPUT were missing from RDC. Additionally,
these were handled as "pseudo" events, but this is not
necessary.

Change-Id: I3478365ac0d78f60a7b63235bea484f3edb8bd16
2021-01-29 14:56:46 -05:00
Chris Freehill 7cf47fb5c9 Add to/correct handling of RDC_EVNT_XGMI_*_THRPUT events
RDC_EVNT_XGMI_[2-5]_THRPUT were missing from RDC. Additionally,
these were handled as "pseudo" events, but this is not
necessary.

Change-Id: I3478365ac0d78f60a7b63235bea484f3edb8bd16


[ROCm/rdc commit: a9d0e037b5]
2021-01-29 14:56:46 -05:00
Chris Freehill e202097fb0 Don't use hwmon# as indicator of gpu
Previously, during the rsmi_init discovery process, the existence
of an hwmon# directory was used to distinguish between gpus nodes
and non-gpu nodes. This isn't reliable in some scenarios. Instead,
the existence of the vbios_version file is used as an
indicator that the node is indeed a gpu.

Change-Id: Icfbe5c42ed0970077b05f25c3d209308a31bec85


[ROCm/rocm_smi_lib commit: ff9546aa62]
2021-01-29 13:05:10 -05:00
Chris Freehill ff9546aa62 Don't use hwmon# as indicator of gpu
Previously, during the rsmi_init discovery process, the existence
of an hwmon# directory was used to distinguish between gpus nodes
and non-gpu nodes. This isn't reliable in some scenarios. Instead,
the existence of the vbios_version file is used as an
indicator that the node is indeed a gpu.

Change-Id: Icfbe5c42ed0970077b05f25c3d209308a31bec85
2021-01-29 13:05:10 -05:00
Chris Freehill fff19b1b3e Don't use hwmon# as indicator of gpu
Previously, during the rsmi_init discovery process, the existence
of an hwmon# directory was used to distinguish between gpus nodes
and non-gpu nodes. This isn't reliable in some scenarios. Instead,
the existence of the vbios_version file is used as an
indicator that the node is indeed a gpu.

Change-Id: Icfbe5c42ed0970077b05f25c3d209308a31bec85


[ROCm/amdsmi commit: ff9546aa62]
2021-01-29 13:05:10 -05:00
Evgeny c5aae30d08 adding gfx1030 blocks
Change-Id: Ide2576939c5321dbe928183a8d9984d5ef87a61b
2021-01-29 08:50:10 -06:00
Evgeny 95ee562f1a adding gfx1030 blocks
Change-Id: Ide2576939c5321dbe928183a8d9984d5ef87a61b


[ROCm/ROCR-Runtime commit: c5aae30d08]
2021-01-29 08:50:10 -06:00
Ori Messinger 46c94662b7 libhsakmt: Add Missing gfx1030 DID
The purpose of this patch is to add a missing device ID for gfx1030.
The missing ID "0x73A1" is now added to the "topology.c" file.

Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: I05a8a55e2c46f941a039fa72a6a5e76bf2a52736
2021-01-29 07:18:49 -05:00
Ori Messinger 57cbb4a5f2 libhsakmt: Add Missing gfx1030 DID
The purpose of this patch is to add a missing device ID for gfx1030.
The missing ID "0x73A1" is now added to the "topology.c" file.

Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: I05a8a55e2c46f941a039fa72a6a5e76bf2a52736


[ROCm/ROCR-Runtime commit: 46c94662b7]
2021-01-29 07:18:49 -05:00
Chauncey Hui 5cf28c4843 SWDEV-2 - Change OpenCL version number from 3249 to 3250
[ROCm/clr commit: b7f0fd0354]
2021-01-29 03:00:06 -05:00
Chauncey Hui b7f0fd0354 SWDEV-2 - Change OpenCL version number from 3249 to 3250 2021-01-29 03:00:06 -05:00
agodavar f379ff60bd SWDEV-269085 - Add support for build with ADDRESS_SANITIZER
Change-Id: I102cbe197727ec059d3dc9fae3a41bb559b0dff1


[ROCm/clr commit: eb58a0285c]
2021-01-29 01:54:43 -05:00
agodavar eb58a0285c SWDEV-269085 - Add support for build with ADDRESS_SANITIZER
Change-Id: I102cbe197727ec059d3dc9fae3a41bb559b0dff1
2021-01-29 01:54:43 -05:00
agodavar e3d44c8508 SWDEV-269085 - Add support for build with ADDRESS_SANITIZER
Change-Id: I102cbe197727ec059d3dc9fae3a41bb559b0dff1


[ROCm/hip commit: e236ae7ea5]
2021-01-29 01:54:43 -05:00
agodavar e236ae7ea5 SWDEV-269085 - Add support for build with ADDRESS_SANITIZER
Change-Id: I102cbe197727ec059d3dc9fae3a41bb559b0dff1
2021-01-29 01:54:43 -05:00
Gilbert Lee 6bf9b0d36a Removing in-place tests from Combined calls (no support for send/recv)
[ROCm/rccl commit: 01a998b17c]
2021-01-28 20:09:03 +00:00
Gilbert Lee 01a998b17c Removing in-place tests from Combined calls (no support for send/recv) 2021-01-28 20:09:03 +00:00
Saleel Kudchadker 7a08212ce1 SWDEV-257787 - Add log for tracking copy signals
Change-Id: I713e8463916a85a634a1ec2309bbd46a11c461a8


[ROCm/clr commit: 629a2d8ef3]
2021-01-28 13:25:49 -05:00
Saleel Kudchadker 629a2d8ef3 SWDEV-257787 - Add log for tracking copy signals
Change-Id: I713e8463916a85a634a1ec2309bbd46a11c461a8
2021-01-28 13:25:49 -05:00
Sarbojit Sarkar 962dff6a61 SWDEV-265469 - added constexpr constructor for dim3
Change-Id: I2e23a403f272d1ff0ade24fec11190d2f3244e5f


[ROCm/clr commit: cee3e1aa37]
2021-01-28 13:23:51 -05:00
Sarbojit Sarkar cee3e1aa37 SWDEV-265469 - added constexpr constructor for dim3
Change-Id: I2e23a403f272d1ff0ade24fec11190d2f3244e5f
2021-01-28 13:23:51 -05:00
Sarbojit Sarkar 3fa56865cc SWDEV-265469 - added constexpr constructor for dim3
Change-Id: I2e23a403f272d1ff0ade24fec11190d2f3244e5f


[ROCm/hip commit: 09b1eb8c9f]
2021-01-28 13:23:51 -05:00
Sarbojit Sarkar 09b1eb8c9f SWDEV-265469 - added constexpr constructor for dim3
Change-Id: I2e23a403f272d1ff0ade24fec11190d2f3244e5f
2021-01-28 13:23:51 -05:00
Vladislav Sytchenko 95655fd5fd SWDEV-268186 - [PAL] Enable MALL for persistent allocations
Change-Id: Ibc21255d5be52289cb4de3de7d12e0415fd1bc91


[ROCm/clr commit: 0d0739a39b]
2021-01-28 12:47:18 -05:00
Vladislav Sytchenko 0d0739a39b SWDEV-268186 - [PAL] Enable MALL for persistent allocations
Change-Id: Ibc21255d5be52289cb4de3de7d12e0415fd1bc91
2021-01-28 12:47:18 -05:00
c981e76efe Clique kernel support (#295) (#15)
* Adding experimental clique-based kernels (opt-in only)

Co-authored-by: Stanley Tsang <stanley.tsang@amd.com>
Co-authored-by: Gilbert Lee <gilbert.lee@amd.com>
Co-authored-by: Wenkai Du <43822138+wenkaidu@users.noreply.github.com>

Co-authored-by: Stanley Tsang <stanley.tsang@amd.com>
Co-authored-by: Wenkai Du <43822138+wenkaidu@users.noreply.github.com>

[ROCm/rccl commit: 3e62ceddc5]
2021-01-28 09:45:01 -07:00
3e62ceddc5 Clique kernel support (#295) (#15)
* Adding experimental clique-based kernels (opt-in only)

Co-authored-by: Stanley Tsang <stanley.tsang@amd.com>
Co-authored-by: Gilbert Lee <gilbert.lee@amd.com>
Co-authored-by: Wenkai Du <43822138+wenkaidu@users.noreply.github.com>

Co-authored-by: Stanley Tsang <stanley.tsang@amd.com>
Co-authored-by: Wenkai Du <43822138+wenkaidu@users.noreply.github.com>
2021-01-28 09:45:01 -07:00
Vladislav Sytchenko b8ce1cc059 SWDEV-1 - Update ISA table
Disable gfx9+ for GSL.

Enable gfx9+ (except MI100/200/300) for PAL.

Change-Id: Ic5a238af939e1d8795503aae39df6a6ece50a9d2


[ROCm/clr commit: 81e331dc42]
2021-01-28 11:26:22 -05:00
Vladislav Sytchenko 81e331dc42 SWDEV-1 - Update ISA table
Disable gfx9+ for GSL.

Enable gfx9+ (except MI100/200/300) for PAL.

Change-Id: Ic5a238af939e1d8795503aae39df6a6ece50a9d2
2021-01-28 11:26:22 -05:00
Vladislav Sytchenko 4987cc7052 SWDEV-1 - Fix Windows build
Change-Id: I20c727a8718941694da6dc024efdfd1eb6af828b


[ROCm/clr commit: f174d98893]
2021-01-28 10:22:25 -05:00
Vladislav Sytchenko f174d98893 SWDEV-1 - Fix Windows build
Change-Id: I20c727a8718941694da6dc024efdfd1eb6af828b
2021-01-28 10:22:25 -05:00
Vladislav Sytchenko e7b6c05a3c SWDEV-1 - Fix Windows build
Change-Id: I20c727a8718941694da6dc024efdfd1eb6af828b


[ROCm/hip commit: ff4b9ab9bc]
2021-01-28 10:22:25 -05:00
Vladislav Sytchenko ff4b9ab9bc SWDEV-1 - Fix Windows build
Change-Id: I20c727a8718941694da6dc024efdfd1eb6af828b
2021-01-28 10:22:25 -05:00
kjayapra-amd 4693ecf345 SWDEV-270013 - Adding code to differentiate between kern_arg + fine_grain and fine_grain segment of system memory.
Change-Id: Id0404b2dab8561946ad0a90a77030582f7e021a4


[ROCm/clr commit: 85fafb0561]
2021-01-28 09:48:17 -05:00
kjayapra-amd 85fafb0561 SWDEV-270013 - Adding code to differentiate between kern_arg + fine_grain and fine_grain segment of system memory.
Change-Id: Id0404b2dab8561946ad0a90a77030582f7e021a4
2021-01-28 09:48:17 -05:00
kjayapra-amd 552218f005 SWDEV-261215 - Updating file handle APIs with file sharing option to accomodate back to back hipModuleLoad calls.
Change-Id: I8fd833b678ca38c60e2d7d02295fd01df0af5592


[ROCm/clr commit: e8208de6c1]
2021-01-28 09:47:41 -05:00
kjayapra-amd e8208de6c1 SWDEV-261215 - Updating file handle APIs with file sharing option to accomodate back to back hipModuleLoad calls.
Change-Id: I8fd833b678ca38c60e2d7d02295fd01df0af5592
2021-01-28 09:47:41 -05:00
kjayapra-amd 81c1c29bc4 SWDEV-270007 - Fixing the hipMemoryAllocateCoherent test binary path for hipMemoryAllocateCoherentDriver test path.
Change-Id: Iae5363685d10fce32ff80ac5d6bfba460dc2b574


[ROCm/clr commit: d13124d90f]
2021-01-28 09:47:01 -05:00
kjayapra-amd d13124d90f SWDEV-270007 - Fixing the hipMemoryAllocateCoherent test binary path for hipMemoryAllocateCoherentDriver test path.
Change-Id: Iae5363685d10fce32ff80ac5d6bfba460dc2b574
2021-01-28 09:47:01 -05:00
kjayapra-amd c01ec4e507 SWDEV-270007 - Fixing the hipMemoryAllocateCoherent test binary path for hipMemoryAllocateCoherentDriver test path.
Change-Id: Iae5363685d10fce32ff80ac5d6bfba460dc2b574


[ROCm/hip commit: f78522c226]
2021-01-28 09:47:01 -05:00
kjayapra-amd f78522c226 SWDEV-270007 - Fixing the hipMemoryAllocateCoherent test binary path for hipMemoryAllocateCoherentDriver test path.
Change-Id: Iae5363685d10fce32ff80ac5d6bfba460dc2b574
2021-01-28 09:47:01 -05:00
Bill(Shuzhou) Liu b8746f7fc0 Disable Autoprov when build rpm
The rdc may install third party library in its local folder. We
need to disable the Autoprov to prevent the rpm add provides
for them.

Change-Id: I18d008c2ca2ecb6bba64467d78b1f8c3a6585aea
2021-01-28 09:34:32 -05:00
Bill(Shuzhou) Liu 8a4a3124f1 Disable Autoprov when build rpm
The rdc may install third party library in its local folder. We
need to disable the Autoprov to prevent the rpm add provides
for them.

Change-Id: I18d008c2ca2ecb6bba64467d78b1f8c3a6585aea


[ROCm/rdc commit: b8746f7fc0]
2021-01-28 09:34:32 -05:00
Satyanvesh Dittakavi 9349ef2f10 SWDEV-269784 - managed memory support for HIP CUDA
Change-Id: I01f9fc64573f402031eceab24395e5cbd93007f9


[ROCm/hipother commit: eeaf35eda3]
2021-01-28 07:51:52 -05:00