커밋 그래프

2800 커밋

작성자 SHA1 메시지 날짜
Ori Messinger 805eeffa32 kfdtest: Fix String NULL Check
MCPU const char * always returns true, so check the value instead.

Before: if (!MCPU) {
After: if (!*MCPU) {

Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: I414e091ca764095937311648c534351d6abf30e6


[ROCm/ROCR-Runtime commit: 5f117f7608]
2023-09-08 16:36:01 -04:00
Jonathan Kim 6746de9752 kfdtest: temporarily exclude debug suspend queues test
For some reason, non-Ubuntu builds have some sort of memory
corruption when running this test, which affect subsequent running
tests.  Disable it for now.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I5f54ee4c63286a33c6948bc818aa1501c4a6751e


[ROCm/ROCR-Runtime commit: 6ec529fe68]
2023-09-08 12:12:13 -04:00
David Yat Sin 2a2555dd52 Update blit shaders for gfx94x
Change-Id: Ic8def71aa0c6ab9a9a758877a65ca6b5625e8f1e


[ROCm/ROCR-Runtime commit: 6ce1586def]
2023-09-08 09:43:31 -04:00
Shweta Khatri e2c5ecb8dc Use LLVM compiler to build blit shaders
Generates shader bytecode stream in amd_blit_shaders_v2.h at build time

Change-Id: I5228ec5442a78d074fd85ca9cd7f7a156dd84da3


[ROCm/ROCR-Runtime commit: 4e675ce730]
2023-09-08 09:42:29 -04:00
David Yat Sin 590cac0321 Fix clang compile warnings
Change-Id: Iea9afc3d998a6c5db28af6c7b54939960b11ae95


[ROCm/ROCR-Runtime commit: 3ee6c9b0e2]
2023-09-07 12:00:02 -04:00
David Yat Sin 3e286607ca Fix for always returning 64 for cacheline size
Change-Id: I0e31d306a2e051ecb9ac019c4e6f5efa25eabba0


[ROCm/ROCR-Runtime commit: 4770b210f6]
2023-08-31 13:50:49 +00:00
David Yat Sin 5b9dcfd0d8 Update interface version for virtual memory APIs
Change-Id: Ifbf1af08ee7aa4d55387ff9786f6a61b89b56f88


[ROCm/ROCR-Runtime commit: 1e7b078628]
2023-08-30 17:01:13 -04:00
David Yat Sin 4e46eded66 Increment HSA API table stepping on new APIs
Add compile time asserts to force incrementing API table STEP versions
each time a new function is added to each table. This is required for
profiler team to be able to add preprocessor macros to determine which
versions contain the new APIs.

Also incrementing the major versions to 2 to indicate new numbering
scheme.

Change-Id: I148a436a5ceab6be3906f8263b40ea9b07841577


[ROCm/ROCR-Runtime commit: 03f2f69d16]
2023-08-29 21:59:36 +00:00
Jonathan Kim c7942fd93f kfdtest: replace 0 initialized dbg structs with memset
Use memset to avoid general 0 set padding issues and ASAN compile issues
for debug tests.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I0a5aca5b7b631083599573b47f1ae87d5d0d5d71


[ROCm/ROCR-Runtime commit: f9e20c8a93]
2023-08-29 11:25:56 -04:00
Lang Yu 9bfa5c5737 kfdtest: add blacklist for gfx1150 and gfx1151
Change-Id: If78840e57c2523696c620d28f4c4ffb004128c0c
Signed-off-by: Lang Yu <Lang.Yu@amd.com>


[ROCm/ROCR-Runtime commit: 65ca3317f2]
2023-08-24 17:27:04 +08:00
Jonathan Kim 9e533f6664 Submit a minimum of 64 DWORDs for SDMA submissions for some GFX9 devices
Some GFX9 devices will drop commands if ring buffer submission is less
than 64 DWORDs.  Pad submission with a NOP head an trailing null
DWORDs in this case.

Change-Id: I850af490fb699f7efe8aef96d97c600a8e76516b


[ROCm/ROCR-Runtime commit: cdd0728d9b]
2023-08-23 13:36:29 -04:00
David Yat Sin 0637810752 Fix memory pool ALLOC_REC_GRANULE query
Also changed enum value to leave gap between enums that only exist in
hsa_region_info_t and enums that exist in both hsa_amd_memory_pool_info_t

Change-Id: I8f9f31200de66648e9328e4203ab283068c993f0


[ROCm/ROCR-Runtime commit: 4317f8dece]
2023-08-22 17:46:48 -04:00
David Yat Sin 777df5c6dc Fix flags passed to thunk for address reserve
Fix flags passed to thunk when reserving address only

Change-Id: Ic91d4c3393cc6a2b98e6bc5ed3575d40fa5e1424


[ROCm/ROCR-Runtime commit: 7be305b83c]
2023-08-22 14:01:49 -04:00
Jonathan Kim ad613e1644 Clean up SDMA ganging
We don't need to keep track of specific blit engines in gang for
submission anymore as ganging early exits on pending bytes.
So tidy up the fluff.

Change-Id: I77e80bf1ad8f561a03fff77bce33aa09d02760c6


[ROCm/ROCR-Runtime commit: 132815bcfb]
2023-08-22 05:57:04 -04:00
Ranjith Ramakrishnan 245c5e2a40 Use memset for initializing variable sized array
In ASAN builds, the compiler used is clang. The initialization of
variable sized array using assignment operator is causing compilation
failure in ASAN builds. Used memset to fix the same.

Change-Id: I02aef3b99a6cad0cce3a378210a48732e07a88fb


[ROCm/ROCR-Runtime commit: 65911e8368]
2023-08-21 12:01:58 -07:00
Jonathan Kim 9bcc4bafdf kfdtest: add trap on wave start and end test
Add test to catch trap on wave start or end override event.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: Icb57af64475fbd2d8a6c0af9a2ee5db5d1a169c6


[ROCm/ROCR-Runtime commit: a3f8085025]
2023-08-18 12:15:08 -04:00
Jonathan Kim 1cd6019517 kfdtest: add address watch test
Address watch test will test read and write operations.
Test will also check if operation is precise if precise
address watch is available.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I7ef835790e26bf6345682755d7dd26a35853bcd5


[ROCm/ROCR-Runtime commit: 8311ca5bfa]
2023-08-18 12:15:07 -04:00
Jonathan Kim 8d3881c268 kfdtest: Add ops for address watch test
Add wave launch override, set/clear address watch and precise memops
test.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: Ib405d5570cd304e02c2e76eca3593cbd9a5937d9


[ROCm/ROCR-Runtime commit: 431dc8d403]
2023-08-18 12:11:48 -04:00
Jonathan Kim 616f5ed0af kfdtest: add memory violation test
Add memory violation detection test.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I1b56f684682836fc84fbec713bd81c53bdd6d413


[ROCm/ROCR-Runtime commit: d4029a9492]
2023-08-18 12:11:48 -04:00
Jonathan Kim e8ea199d97 kfdtest: allow toggle of dispatch privilege
For GFX11 debugger testing, waves require to start in non-priv mode for
some test cases, so allow tester to set this.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: Iee93fda926bfd336d51c79c086f1f75bc35b70e5


[ROCm/ROCR-Runtime commit: 6c5121faff]
2023-08-18 12:09:07 -04:00
Jonathan Kim 704d9c5e19 Fix SDMA ganging circular deadlock in oversubscription
When oversubscribing SDMA gangs, a circular deadlock can occur since
gang enqueue is staggered with respect to SDMA engine leader based
on source to destination.
As a result, an enqueued leader may be waiting on a gang item that is
waiting on another enqueued leader or gang item and so on.

To prevent this, first lock the submission to ensure dma status query
and submissions are atomic.  Once this is in place, be more stringent
with ganging in that all SDMA engines must be available in order to gang.

Finally, re-enable SDMA ganging by default.

Change-Id: I4511e3487db9d26475b5aece4897f10168cc5322


[ROCm/ROCR-Runtime commit: 8f21793a3e]
2023-08-17 08:49:09 -04:00
Jonathan Kim 58d5f7354f Update D2D SDMA ganging for non-SPX modes
xGMI for compute partitioning in non-SPX modes does not have
a reported bandwith.
Fix it to at most 2 since each partition is either bounded
by the number of xGMI links or the number of available
SDMA contexts.

Change-Id: I09094bd7548d9eee6f039b0efe849838e5de166e


[ROCm/ROCR-Runtime commit: 4c74e47e91]
2023-08-17 07:25:08 -04:00
Jonathan Kim 2994cfa875 Bump the number of SDMA engines for gfx940
GFX940 can support up to 16 SDMA engines so bump it.

Change-Id: I41a95e66383036735712e317a57b239d84fcb78d


[ROCm/ROCR-Runtime commit: 30982ff6aa]
2023-08-17 07:25:08 -04:00
Jonathan Kim 64e0037743 Break when finding ganged agent
There's no need to keep looking in the list once we find a ganged agent.

Change-Id: Ia0b9b484c88221a7966a814456942c19b1741978


[ROCm/ROCR-Runtime commit: f8664e88e0]
2023-08-17 07:25:08 -04:00
David Yat Sin 38bec00960 Temporarily disable SDMA ganging by default
SDMA ganging is causing some regressions with some applications hanging.
Temporarily disabling SDMA ganging by default until issue is fixed.

Change-Id: I65e172923a53a967df27b30d969ad5d215c4fa09


[ROCm/ROCR-Runtime commit: a20a0a5bac]
2023-08-15 23:17:34 +00:00
David Yat Sin 5564790017 Revert "Adding documentation for SDMA environment var"
This reverts commit 56ccf828bc.

Replaced by commit 3b3f14c06e8a2fab717f0b82aba3c72d74bb9574.

Environment variables documented in:docs/environment_variables.md

Change-Id: I8da0d971eb98554b4bd1b884617a439f1b20ed5b


[ROCm/ROCR-Runtime commit: 93401e3c8c]
2023-08-10 09:55:42 -04:00
Ranjith Ramakrishnan 78f32a3b5e Disable file reorg backward compatibility support by default
Change-Id: I157e05e52a1a61b86fa2fc6f29d31361a688fa10


[ROCm/ROCR-Runtime commit: 96c8bb11b1]
2023-08-09 12:31:27 -04:00
Jonathan Kim 526952d090 kfdtest: add suspend resume test
Add queue suspend and resume test.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I2ade721026cbb458a3597b7858a164e70fe05f4f


[ROCm/ROCR-Runtime commit: d20f0bbb90]
2023-08-09 09:26:46 -04:00
Jonathan Kim f404315977 kfdtest: add snapshot operations
Add queue and devices snapshot operations.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I836884c9f3b65dd9e5e444d554d3eb87938e1634


[ROCm/ROCR-Runtime commit: b0e84183c1]
2023-08-09 09:26:29 -04:00
Jonathan Kim c279ff343f kfdtest: add suspend and resume queues operation
Add base debug operations to suspend and resume queues.
Routine will return the number of queues successfully
suspended or resumed.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I8f18317f70464b04231c5cf822e11d545ebfa02a


[ROCm/ROCR-Runtime commit: 5a675921ea]
2023-08-09 09:26:09 -04:00
Jonathan Kim a10eaed8dc kfdtest: add hit trap event test
Check that a jump to trap event can be picked up by the debugger.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: Iad5f87092f2b82d5018013bba548979122a9bd02


[ROCm/ROCR-Runtime commit: b77189cf83]
2023-08-08 16:01:23 -04:00
Jonathan Kim 3bddb35c78 kfdtest: add set exceptions enable base debug operation
Add set exceptions enabled debug option

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I6ee1769bbbb90a74074d8100974c4bfeabaf7f2c


[ROCm/ROCR-Runtime commit: 97fc25bb8d]
2023-08-08 16:01:03 -04:00
Jonathan Kim 115ff68bda kfdtest: add runtime enable and attach test
Add debug attach and runtime enable test for attaching to a spawned and
running process.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I72302ff73494d9dae0c79a299508085d7ca0552b


[ROCm/ROCR-Runtime commit: 097ee967d1]
2023-08-08 16:00:44 -04:00
Jonathan Kim fb140303bb kfdtest: add query operations
Add polling query debug event operation.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: Ic82ce4a393bfb28c9f32e7920f80c12da7f627d5


[ROCm/ROCR-Runtime commit: bfb0d15ee8]
2023-08-08 16:00:25 -04:00
Jonathan Kim 4696060070 kfdtest: add send exception operation
Add send exception operation.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: Iecb43004a1a7ffc75e94badf203cd0927ffe0909


[ROCm/ROCR-Runtime commit: c7129edcb8]
2023-08-08 16:00:04 -04:00
Jonathan Kim 1435bb4887 kfdtest: add base debug class and debug attach/detach operation
Add base debug class and attach/detach operations.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I60f3c166646f05838fec208ac2f59bba998c63f8


[ROCm/ROCR-Runtime commit: dd56b38c2f]
2023-08-08 15:59:35 -04:00
Ranjith Ramakrishnan 670217a201 Disable file reorg backward compatibility support by default
Change-Id: Ib53a4d0476ec598025d4f1f98414e0e425bb0e49


[ROCm/ROCR-Runtime commit: bb4756d2e0]
2023-08-07 09:38:12 -07:00
David Yat Sin b5f6c688b6 Keep libdrm device_handle on older libdrm
Even if the version of libdrm older and does not support the
amdgpu_device_get_fd function, the device_handle stored in
amdgpu_handle[] is still valid and can be returned via
hsaKmtGetAMDGPUDeviceHandle.

Change-Id: I024a3e82e6cfebac5577aefe359b067746c4023e


[ROCm/ROCR-Runtime commit: 66b66e42cd]
2023-08-01 10:52:26 -04:00
David Yat Sin 47b40068f6 Fix compile error when using clang
Fix compile error due to arithmetic on void*
Fix some compile warnings

Change-Id: I03ded438c5af77ba61c0a7017be5d4fe1e16c16c


[ROCm/ROCR-Runtime commit: 93aff0b439]
2023-07-31 18:29:19 +00:00
Jonathan Kim ae3b48d227 Enable D2D SDMA Ganging over xGMI
Use all available SDMA engines capped by xGMI bandwith for
all D2D copies within a hive.

By default, set the latency boundary copy size as 4KB and below.
Any copy size in within this boundary will not gang.

Avoid oversubscribing engines by not ganging on engines with
pending non-ganged work.

An enviroment variable HSA_ENABLE_SDMA_GANG has been provided
to override default ganging behaviour.

Change-Id: Iccde76aa1af1d47ea2a151789432c9db4f0ffa8d


[ROCm/ROCR-Runtime commit: 7df0167821]
2023-07-27 08:58:26 -04:00
Jonathan Kim 8d0be0c17f Silence parenthesis warnings in mem API
Fix KFD version checking parenthesis warnings on compile.

Change-Id: I89c46ea84a8d75b761d8c40ff62d008c7afbef2d


[ROCm/ROCR-Runtime commit: c5dbb93e59]
2023-07-26 16:14:40 -04:00
David Yat Sin a8e34eaec8 Revert "Add support for GC 11.5.0 and 11.5.1"
Reverting this as current mainline compiler branch does not support
gfx1150/gfx1151 yet. Will bring back later.

This reverts commit 75ce1848cf.

Change-Id: I31ff4fb2d5817538094a7ffaeba96dd6a7d660c7


[ROCm/ROCR-Runtime commit: ebc51dd0eb]
2023-07-26 15:03:54 +00:00
Jonathan Kim 72bd0f3a8d libhsakmt: add debug trap thunk call for testing
Add generic thunk call for debug testing that assumes
caller populations trap arguments correctly.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I33a0bc66ca77e29f5b663d4bfe73f8684df8bfb6


[ROCm/ROCR-Runtime commit: aaab019960]
2023-07-26 10:29:27 -04:00
Jonathan Kim 6170aadb8c kfdtest: remove deprecated debug references
Remove all unused material from KFDDBGTest.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I13ed68656efadef7bbaf8bb737ce5a04829eca9b


[ROCm/ROCR-Runtime commit: 98c6784cc1]
2023-07-26 10:29:23 -04:00
Jonathan Kim 4641eeebfe libhsakmt: remove old debugger versioning
Current debugger uses KFD version directly.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Change-Id: I212a53560a94dd24c599addce72f59c527c8af25


[ROCm/ROCR-Runtime commit: 8471f80bac]
2023-07-26 09:41:38 -04:00
Philip Yang d92f10f8e8 kfdtest: KFDSVMEvictTest support large VRAM or small system memory
For xnack off, skip SVM evict tests if memory allocation size is larger
than 15/16 total system memory, because the test may fail to allocate
CWSR svm range to create queue after allocating test memory.

Limit eviction size from total VRAM size to 1/2 total VRAM size,
because for 192GB VRAM, evict 192GB may takes more than 120 seconds
and cause test timeout failed.

Change-Id: Ib1483b9aab580a8539187b2943cadea0fd5a7c71
Signed-off-by: Philip Yang <Philip.Yang@amd.com>


[ROCm/ROCR-Runtime commit: a395dd7306]
2023-07-25 11:11:55 -04:00
David Yat Sin 2cfb82a1a6 Add agent query for nearest CPU agent
Add agent info query to return nearest CPU agent. This can be used to
determine which CPU agent is in the same NUMA region as the GPU agent.

Change-Id: I5400b4347ffbf4d2a836df31c4de443a38b0ecd1


[ROCm/ROCR-Runtime commit: 469defa78a]
2023-07-24 13:59:13 -04:00
Jonathan Kim 8e16b26347 Silence implicity conversion warnings in exception handling
Silence unnamed enum warning in error code comparison

Change-Id: I008b269c106bbad83a1f7588e7b4ec89ec17d37d


[ROCm/ROCR-Runtime commit: 0d14144e3a]
2023-07-24 10:06:55 -04:00
Jonathan Kim a29b3ab868 Fix out of order initializer for memory region
Silence out of order initializer compile warnings during memory region
initialization.

Change-Id: Idbbdd93d3ea8cda289d25a473b3882b920b2e8d8


[ROCm/ROCR-Runtime commit: 42274cfc59]
2023-07-24 09:58:37 -04:00
Lang Yu 75ce1848cf Add support for GC 11.5.0 and 11.5.1
Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Change-Id: I3c4116e78a5c1ddac2389f5fece57485bdb17f68


[ROCm/ROCR-Runtime commit: e877840197]
2023-07-22 16:06:22 +08:00