Commit grafiek

763 Commits

Auteur SHA1 Bericht Datum
Kent Russell 08bded0cb3 kfdtest: Add sanitizer flags after C flags are set
Otherwise they get overwritten

Change-Id: I9042422d4515e7ac812ed34779906b0b2c44545c


[ROCm/ROCR-Runtime commit: 92ad039915]
2020-12-24 10:32:01 -08:00
Kent Russell 17a83e8e69 Remove address-sanitizer debug messages
Change-Id: I08509aaed36459329f0a65264e42f287c27f4a18


[ROCm/ROCR-Runtime commit: f6f47aa43d]
2020-12-24 10:32:01 -08:00
Kent Russell 8a95a2d584 kfdtest: Support address sanitizer in KFDTest
Change-Id: Iee1182608ddc9896c82feb5004b3fe078d3d3223


[ROCm/ROCR-Runtime commit: 9cca1216e9]
2020-12-24 10:32:01 -08:00
Kent Russell 0682ddc56a Set -no-undefined properly if it's CC
Address-sanitizer doesn't like it at all. And it's called differently
under clang than gcc, so adjust accordingly

Change-Id: Iebe8cd68618d3f7a4c310419c64b4f73d7ecfda4


[ROCm/ROCR-Runtime commit: 8e0a9aa417]
2020-12-24 10:32:01 -08:00
Kent Russell 5b7549331f CMakeLists: Address-sanitizer fix and cleanup
Move all the logic into 1 spot, and make sure -fsanitize=address is also
passed to the library flags

Change-Id: I7b60629d32df6436b5c7ad37997fe14ea48f5d72


[ROCm/ROCR-Runtime commit: 3d9f60d7fe]
2020-12-24 10:32:01 -08:00
Gefei Jiang 4b5fa5bb42 CMakeLists.txt: Address Sanitizer Support
1. add sanitize flag to link flags
  2. use ${ADDRESS_SANITIZER} as a condition to turn on/off
     instead of (DEFINED ADDRESS_SANITIZER).
     The latter will always turn the feature on regardless of
     the value as long as there is "-DADDRESS_SANITIZER in
     cmake command line,which will be an issue when merging to
     the mainline
Amended: put -fsanitize=address at the begining in link flags

Change-Id: I84df0e5b6d7fb8f02f18bf7961f25f15cac10443
Signed-off-by: Gefei Jiang <gefei.jiang@amd.com>


[ROCm/ROCR-Runtime commit: d3bc75d229]
2020-12-24 10:32:01 -08:00
Gefei Jiang e93ba06265 CMakeLists.txt: Address Sanitize Support
ROCMOPS-1249
	correct if statement and -f flag name

Change-Id: I92e9aa30b1c81f855ad269c0c686ec1e136a85fd
Signed-off-by: Gefei Jiang <gefei.jiang@amd.com>


[ROCm/ROCR-Runtime commit: b92d28bd71]
2020-12-24 10:32:01 -08:00
Gefei Jiang 8d6f41350a CMakeList.txt -- Support Address Sanitize
ROCMOPS-1249
	append address sanitize flag

Change-Id: Ie5d1e5b8b93022b80e0ca74106a16d53d52e41af
Signed-off-by: Gefei Jiang <gefei.jiang@amd.com>


[ROCm/ROCR-Runtime commit: f0e6e7ae17]
2020-12-24 10:32:01 -08:00
Yifan Zhang 72b5ce407a kfdtest: Take vram size into account when calculate buffer number.
Vram size is relatively smaller in APU, e.g. 512MB.
Current MMBench doesn't support small vram system.
Running MMBench may have below errors:

[ RUN      ] KFDMemoryTest.MMBench
[          ] Found VRAM of 512MB.
[          ] Test (avg. ns)        alloc   mapOne  umapOne   mapAll  umapAll     free
[          ] --------------------------------------------------------------------------
[          ]   4K-SysMem-noSDMA         4569    20098     1292    18835      926     2218
[          ]  64K-SysMem-noSDMA        12738    20469     1030    19201     1293     4560
[          ]   2M-SysMem-noSDMA       256384    21020     1022    20568     1196    36294
[          ]  32M-SysMem-noSDMA      4031812    83750     5406    61156     4312   535656
[          ]   1G-SysMem-noSDMA    129260000   427000    34000   390000    30000 18548000
[          ] --------------------------------------------------------------------------
[          ]   4K-VRAM-noSDMA         3594    19637      979    19624     1357     2829
[          ]  64K-VRAM-noSDMA         3540    21062     1407    19614     1654     3024
/home/foreman/build/hsakmt-roct-amdgpu-1.0.9/sources/libhsakmt/tests/kfdtest/src/KFDMemoryTest.cpp:1119: Failure
Value of: (hsaKmtAllocMemory(allocNode, bufSize, memFlags, &bufs[i]))
  Actual: 6
Expected: HSAKMT_STATUS_SUCCESS
Which is: 0
[  FAILED  ] KFDMemoryTest.MMBench (723 ms)

Fix this issue by changing buffer number calculation in MMBench.

Change-Id: I5cce95707a048248f1e825c807586818619eddaf
Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>


[ROCm/ROCR-Runtime commit: 742f718722]
2020-12-17 07:41:24 -05:00
Chengming Gui 5d08071b4f kfdtest: remove unsupported modifier 'offset'
fix 
v2: fix VGPR conflict
v3: use s_addc_u32 to replace s_add_u32

Signed-off-by: Chengming Gui <Jack.Gui@amd.com>
Change-Id: I8fe6bf1f5bf99544038ad16128c2bebd559d3da9


[ROCm/ROCR-Runtime commit: 3ed8b96bf0]
2020-12-14 17:29:13 +08:00
Huang Rui d72d4943b4 libhsakmt: add gfx1033 support
This patch is to add Van Gogh support on thunk.

Change-Id: I75819329b865e4c38c097e83e3a0cb4e4f566fa2
Signed-off-by: Huang Rui <ray.huang@amd.com>


[ROCm/ROCR-Runtime commit: 9600760ff7]
2020-12-08 23:54:46 -05:00
changzhu 2608314fae Add distinguish for iommuv2/dgpu_fallback when getting gpuName
The memory tests between iommuv2 and dgpu_fallback are different.So it
needs to ditinguish them.

Change-Id: Icc64e9ae0fc1638c3d148795a5f247d9e5e8e503
Signed-off-by: changzhu <Changfeng.Zhu@amd.com>


[ROCm/ROCR-Runtime commit: 39386c03bf]
2020-12-04 02:24:49 -05:00
Philip Cox c03f46e198 kfdtest: increase default timeout to 10,000
The default kfdtest timeout is not enough for certain platforms, and
tests are failing.

Change-Id: I2027eadcbeb12a2fbbc9c55f92f31869fa13dbcb
Signed-off-by: Philip Cox <Philip.Cox@amd.com>


[ROCm/ROCR-Runtime commit: 4bbfbe7789]
2020-11-27 15:06:41 -05:00
Gang Ba 3d11816985 kfdtest: check peer accessible with new function
check GPU peer accessible with p2p_links in system

Signed-off-by: Gang Ba <gaba@amd.com>
Change-Id: I026f16564303b687811d6648f0b7f84be6819979


[ROCm/ROCR-Runtime commit: 8e94dde685]
2020-11-26 10:34:06 -05:00
Kent Russell e39dab44e3 Look in /opt/rocm* for SMI for setting clocks to high
Now that symlinks aren't necessarily guaranteed, use "find" to try to
find the rocm-smi, and clarify the error message if it is not found

Also tie in a fix for parsing the output now that the output has changed

Change-Id: I2081442a71731c186c3ad00585a2ba6e8a8e5a28


[ROCm/ROCR-Runtime commit: 2651ce37d8]
2020-11-23 14:05:10 -05:00
Aakash Sudhanwa 0e2e1aca74 Revert "CMakeLists: Fix RPM dependency declaration"
This reverts commit 4aae34276b.

Reason for revert: <INSERT REASONING HERE>

Change-Id: Ieaf2da0067c3e89577569c5d478c70b97ca5f5ca


[ROCm/ROCR-Runtime commit: ac6e96e7a3]
2020-11-17 20:10:45 -05:00
Gang Ba 98bb73932d libhsakmt: Create P2P links
1. Create P2P links
2. Determine FRAMEBUFFER_PUBLIC/PRIVATE only based
   host-accessibility, not peer-accesssibility

Signed-off-by: Gang Ba <gaba@amd.com>
Change-Id: I15fccdc60386b453e2a47849a16df15157324b21


[ROCm/ROCR-Runtime commit: bedecc5957]
2020-11-17 15:43:12 -05:00
Kent Russell 4aae34276b CMakeLists: Fix RPM dependency declaration
RPM needs _REQUIRES at the end, not _DEPENDS, and also requires a space
before the version of the required package.

Change-Id: I9dd70bd92fc2407b7e8b31e4d46df43c52438a65
Signed-off-by: Kent Russell <kent.russell@amd.com>


[ROCm/ROCR-Runtime commit: 089fdeb1fe]
2020-11-17 12:36:05 -05:00
Gang Ba 0ddd455ead libhsakmt: add Streaming Performance Monitors APIs
Signed-off-by: Gang Ba <gaba@amd.com>
Change-Id: Iab9a98fa2079b7cae7158c524479dfc3fa672407


[ROCm/ROCR-Runtime commit: e8c0426c54]
2020-11-16 16:36:21 -05:00
Ashutosh Mishra 5026ed2f8a Standardizing Package names
: Enables standards compliant package naming for debian and rpm

Signed-off-by: Ashutosh Mishra <ashutosh.mishra@amd.com>
Change-Id: I177af15ec7a3f909d05135be30a0acc7b0b20745


[ROCm/ROCR-Runtime commit: aefc173da4]
2020-11-11 00:37:20 -05:00
Tianci.Yin 349fe6a583 kfdtest: add DID for gfx1010 blockchain SKU
Change-Id: Icd52c5db4dd975086fcfb13deb6727919c1f5809
Signed-off-by: Tianci.Yin <tianci.yin@amd.com>


[ROCm/ROCR-Runtime commit: 1231570441]
2020-11-04 01:14:32 -05:00
Chengming Gui 55fe1ba5a3 kfdtest: update shader code for gfx10.3 kfdtest
s_store_* instruction set was retired from gfx10.3

Signed-off-by: Chengming Gui <Jack.Gui@amd.com>
Change-Id: Ibe41a3fe7e053fb345b1af6ad4abc22a0885bc81


[ROCm/ROCR-Runtime commit: f283fe2854]
2020-11-03 22:25:39 -05:00
Felix Kuehling d8ded1fa6e Revert "libhsakmt: optimize system memory allocation"
This reverts commit 35b07e1e28.

Reason for revert: This commit caused a regression rocrtst memory
subtest: Maximum Single Allocation in Memory Pools failed.


Change-Id: I15330625603f893200a08cd8b5b097f9bf95361f


[ROCm/ROCR-Runtime commit: e515fd818b]
2020-11-03 12:19:56 -05:00
Jeremy Newton a803c56889 kfdtest: Use ldflags for drm
This fixes a build issue with kfdtest and the amdgpu pro driver build.

This was requested as kfdtest is needed for regular testing due to the
inclusion of the ROCr/KFD stack in the amdgpu pro driver (OSGSUP-199)

Change-Id: I224d2e9ee3f02065596890b4d8226484f4fac04f
Signed-off-by: Jeremy Newton <Jeremy.Newton@amd.com>


[ROCm/ROCR-Runtime commit: 8026ba250c]
2020-11-02 16:46:45 -05:00
Kent Russell 693a1d2b95 kfdtest: Fix bit compare logic by using | instead of ||
We want to compare bits, not check if a defined value is true

Change-Id: Ie51ede96d18eae01aff6677d852a056ee12bd9c6


[ROCm/ROCR-Runtime commit: e34dfa8ebd]
2020-11-02 06:03:59 -05:00
Kent Russell 9057877de1 kfdtest: Add missing default switch label in GetQueue function
There is no default case, and we were missing a few types defined from
hsakmttypes.h. This was found via clang

Change-Id: I26193cb111a9d8220b1eff21c7313fe060288f36


[ROCm/ROCR-Runtime commit: 761d9d84d2]
2020-11-02 06:03:59 -05:00
Kent Russell 85b0b04903 Remove extra strlen call
While the ternary is nice to read, strlen in general is an expensive
call, so call it once and check if the value is greater than our maximum
allowable string length and adjust accordingly

Signed-off-by: Kent Russell <kent.russell@amd.com>
Change-Id: Id744f2ba0eb81bb2b3c52eb69f38a615398a655d


[ROCm/ROCR-Runtime commit: 025036a662]
2020-10-30 06:44:48 -04:00
Felix Kuehling acdda4811b libhsakmt: handle GPU mapping errors
Don't update the vm_object if GPU mapping failed. Print an error message
to help diagnose underlying problems.

Change-Id: I801ab6fe6c155bd25e6c0358007c106a4a019480
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>


[ROCm/ROCR-Runtime commit: c7e6f5a274]
2020-10-26 18:46:24 -04:00
Felix Kuehling 35b07e1e28 libhsakmt: optimize system memory allocation
Use MAP_POPULATE when allocating anonymous system memory for later
GPU mapping as a userptr. This can speed up large allocations by
more than factor 2. I suspect populating pages in this way is more
efficient than the CPU page fault code path triggered by
get_user_pages in the kernel.

Change-Id: I188bbc1462ccb650d48cbfb1080dbb8eb7ada8b5
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>


[ROCm/ROCR-Runtime commit: 8f26c0c40c]
2020-10-26 18:44:20 -04:00
Laurent Morichetti d933e70089 libhsakmt: Fix the ctrl stack size calculation
On gfx9, the maximum number of wavefronts per queue is the minimum of
40 waves per compute units, or 512 waves per shader engine.  On gfx10,
there can only be 32 waves per compute units.

Signed-off-by: Laurent Morichetti <laurent.morichetti@amd.com>
Change-Id: I148d1a4fe6c07cdbfaa1f77939eb29311c81c008


[ROCm/ROCR-Runtime commit: 783e346777]
2020-10-16 12:35:11 -07:00
Chengming Gui 120457680e kfdtest: Add gfx1032 support
Signed-off-by: Chengming Gui <Jack.Gui@amd.com>
Change-Id: I04cd412a5e243dfe7aa7596287341e1671c1521a


[ROCm/ROCR-Runtime commit: 4a9d55c414]
2020-10-12 11:30:39 +08:00
Chengming Gui c5429d56c1 libhsakmt: Prepare gfx1032 support
PCI IDs have yet to be added later.

Signed-off-by: Chengming Gui <Jack.Gui@amd.com>
Change-Id: I28f657201868423012e856df4310a493b7cd5752


[ROCm/ROCR-Runtime commit: b543f4f77c]
2020-10-12 11:30:39 +08:00
Laurent Morichetti 18240aa6d6 Update the context save area size
Reserve some space in the context save area for the debugger's
use. There should be 32 bytes per wave for a given queue.

Change-Id: I65ddb6123d0f6afd3149844617ad19023009101d


[ROCm/ROCR-Runtime commit: 2ed2e46b9b]
2020-10-05 12:10:58 -07:00
Jay Cornwall 7821213760 libhsakmt: Limit control stack size on gfx10
The queue control stack size cannot exceed 0x7000 on ASICs
gfx1010 through gfx1031. The lower limit is not achievable
with AQL so this should have no practical effect.

Fixes control stack size overflow on large ASICs.

Change-Id: Ib78cf6e4c5f096044bf8de24debe211689891caa
Signed-off-by: Jay Cornwall <jay.cornwall@amd.com>


[ROCm/ROCR-Runtime commit: 44f80d170d]
2020-09-28 18:14:57 -05:00
Harish Kasiviswanathan 0e9dbc61d2 libhsakmt: add device ID for gfx1030
Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Change-Id: I3024bf706f88c0e612391f6d8045020101007bdc


[ROCm/ROCR-Runtime commit: 3311785c7b]
2020-09-21 14:23:46 -04:00
Chengming Gui 90298c1436 kfdtest: fix MTYPE error when init sdma fence packet
Signed-off-by: Chengming Gui <Jack.Gui@amd.com>
Change-Id: I14a1ef204dbd5ab1e9f1840b9555f88b0df361c0


[ROCm/ROCR-Runtime commit: befc7edaea]
2020-09-21 03:57:12 -04:00
Jinzhou.Su 8fc695a7c1 kfdtest: fix gfx90c blacklist issue
This patch is the hot fix to update gfx90c blacklist 

Change-Id: I41f48154ad5ec3035fcb7891a224fc940dca991f


[ROCm/ROCR-Runtime commit: 7efa823ec8]
2020-09-16 09:42:01 +08:00
Jinzhou.Su 0b78c33201 kfdtest:Update gfx90c blacklist
1. Add KFDEvictTest.* for gfx90c based on CI test results

2. Remove SDMA blacklist based on SDMA issue fixed:

Change-Id: I86910fc98a5141f29959b35248a900f0c098a6e8


[ROCm/ROCR-Runtime commit: 36249ddc0e]
2020-09-14 22:49:53 -04:00
Huang Rui 2b2835e596 libhsakmt: fix to update the param number after remove to dgpu input
This patch is the hot fix to fix the param number checking after remove
dgpu input.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Change-Id: Ic980588f78616f99076de742af580afb4273fb2f


[ROCm/ROCR-Runtime commit: 8fc816affe]
2020-09-11 10:25:37 -04:00
Huang Rui e2de06530f libhsakmt: update gfx90c isa version
gfx90c should use GFX902 which is the same with gfx902.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Change-Id: Id24dc2c85c9f49f36b00889c3b8b1b19cce34e09


[ROCm/ROCR-Runtime commit: 8ea0d49337]
2020-09-09 22:10:58 -04:00
Kent Russell 0ec9c67e02 Remove -dev make instructions
These are removed now that we've consolidated the dev package
information into CMakeLists.txt from hsakmt-dev.txt.

Change-Id: I49496ec5def85b0af7fa6b15110910528a8e0be0


[ROCm/ROCR-Runtime commit: 654ee83ac8]
2020-09-09 11:28:18 -04:00
Kent Russell 3262952918 Fix lintian errors in deb packages
Add extended descriptions and e-mail address to CMakeLists
A lintian error will remain regarding stripping the .so, as we
will not be doing this for Release versions of the hsakmt .so

Signed-off-by: Kent Russell <kent.russell@amd.com>
Change-Id: I41c768dee28c0564d92b9c103a6e2d97590e4589


[ROCm/ROCR-Runtime commit: 0a4b23d625]
2020-09-09 11:28:18 -04:00
Kent Russell f28303dd25 README: Update README info and installation instructions
Added caveat regarding Ubuntu 20.04 /dev/kfd permissions
Edit copyright date, remove outdated HW support list

Change-Id: I3138058a01db6c46cf624bcfb8cfb04c164d9133
Signed-off-by: Kent Russell <kent.russell@amd.com>


[ROCm/ROCR-Runtime commit: ae4e980ea8]
2020-09-09 11:28:18 -04:00
Huang Rui bbc6eb6f3f libhsakmt: remove is_dgpu flag in the hsa_gfxip_table
Whether use dgpu path will check the props which exposed from kernel.
We won't need hard code in the ASIC table.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Change-Id: I0c018a26b219914a41197ff36dbec7a75945d452


[ROCm/ROCR-Runtime commit: ad87f38dad]
2020-09-09 20:56:50 +08:00
Huang Rui 51a5e47a1b libhsakmt: implement the method that using flag which exposed by kfd to configure is_dgpu
KFD already implemented the fallback path for APU. Thunk will use flag
which exposed by kfd to configure is_dgpu instead of hardcode before.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Change-Id: I445f6cf668f9484dd06cd9ae1bb3cfe7428ec7eb


[ROCm/ROCR-Runtime commit: 12813691a2]
2020-09-09 20:56:39 +08:00
Oak Zeng a1aec2cc19 CWSR control stack size calculation for gfx10
Gfx10 need 12bytes/wave control stack

Change-Id: I6c6f2819572e6b43aa3140d4dbe79d930e4c1c9c
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Signed-off-by: Jay Cornwall <Jay.Cornwall@amd.com>


[ROCm/ROCR-Runtime commit: 3d3b28b670]
2020-09-01 21:34:00 -07:00
Philip Cox 52a223ac90 libhsakmt: call madvise() from fmm_allocate_device
This is needed to avoid additional references to mapped BOs in child
processes that can prevent freeing memory in the parent process and lead
to out-of-memory conditions.

Change-Id: I25c90510a14dde515cc23ea5dc1f68e8d7e37a66
Signed-off-by: Philip Cox <Philip.Cox@amd.com>


[ROCm/ROCR-Runtime commit: f7a3427c99]
2020-08-19 13:33:47 -04:00
Jinzhou.Su 32cb33a6a6 kfdtest: remove test cases on gfx90c which block IQE test
SDMA_BLACKLIST, KFDMemoryTest.LargestSysBufferTest,
KFDQMTest.BasicCuMaskingEven

Change-Id: Ic603452ede5ef14f5fcb49925d57d3a656d48a08
Signed-off-by: Jinzhou.Su <Jinzhou.Su@amd.com>


[ROCm/ROCR-Runtime commit: a69830917f]
2020-08-19 10:20:24 +08:00
Kent Russell 0c1420ef32 Fix GCC warning regarding strncpy in CPU info
strlen(src) should not be used as the length in strncpy. Use memcpy
since we know the length of the string, and ensure that we
NULL-terminate regardless of length

Signed-off-by: Kent Russell <kent.russell@amd.com>
Change-Id: I21cc6d106510c69464e7ac9d3fc7da3a1e6d1a68


[ROCm/ROCR-Runtime commit: 04f6b9e16b]
2020-08-14 07:10:19 -04:00
Kent Russell 0fd38ef996 Fix typo lager->larger
Signed-off-by: Kent Russell <kent.russell@amd.com>
Change-Id: I188d629b6441e5ebb14f104869e871d003c78c9d


[ROCm/ROCR-Runtime commit: 6085baa2dc]
2020-08-13 06:34:42 -04:00