c3bd229f4f
SWDEV-422207 - Optimize graph end detection
...
- Do not use extra barrier to detect graph end. If its a kernel node we
can use a completion signal for the last packet. Saves roughly 6us for
Phantom testcase per graph launch.
Change-Id: I5e0c2479d9964fbeda86ed97533f6718f49a7f91
2023-11-10 11:57:02 -05:00
f5c6fc4dfa
SWDEV-422207 - Report TS for Accumulate command
...
Change-Id: Iba193a6068c1a2d25c2136643faee2c1e2591a07
2023-11-07 18:19:40 +00:00
40f41f4d0b
SWDEV-422207 - Track commands for capture
...
- Track all captured commands under a new AccumulateCommand
- Add begin() and end() methods to capture commands
- Explicit TS object now passed to certain methods because
profilingBegin() and profilingEnd() now happen separately and thus can
run into threading issues
Change-Id: I171106bdcad72b057836cb2f3fc398db3533119f
2023-11-03 05:09:04 +00:00
1338ff37e8
SWDEV-301667 - Cleanup unused paths
...
- Refactor code and cleanup logic for callback saving for event records
Change-Id: I5c56aa8e9c968a5bca70fb07ad1796da318e9e89
2023-11-02 11:43:41 -04:00
182f6e2e6d
SWDEV-422207 - Fix profiling check
...
Change-Id: Ia6ca28f0b593ea1680732684dd10f40c0043c973
2023-10-25 05:37:55 +00:00
38d2c56784
SWDEV-422207 - Handle nonkernel nodes for graph opt
...
- Support graph with different types of nodes with single
branch when DEBUG_CLR_GRAPH_PACKET_CAPTURE flag is enabled
Change-Id: I149a8629769cd0d5849ffefb04f1352668a685b6
2023-10-24 18:36:06 +00:00
de2b06a7a7
SWDEV-301667 - Reset addSystemScope_
...
If we submit a systemScope Barrier, we should reset this flag as there
is no need for dispatch AQL again to flush caches/HDP
Change-Id: I55710feb4ba6650852e785b5cadfa64c6b9ce14e
2023-10-23 16:53:47 -04:00
2d8c65eedf
SWDEV-427493 - Fixed segfault after accessing deleted timestamp
...
Change-Id: I478008f03e2bb4d9ac1f0865605edbf9f8254fea
2023-10-17 15:24:07 -04:00
3ef829939a
SWDEV-413997 - Initial VMM changes for ROCm path.
...
Change-Id: I4405fd7b53182eb4c4622835c811c0dc08461537
2023-10-16 11:29:16 -04:00
bf8baeecb3
SWDEV-301667 - Track last used SDMA engine per queue
...
- Track last SDMA engine per queue, this results in better scheduling
- Reset last SDMA engine upon batch completion. That ensures we dont get
blocked if the same engine is used by another concurrent copy
Change-Id: Id53111980da7ee41d5c932fb44e4aab5b1e065a3
2023-10-10 12:13:11 -04:00
6a8bc3c718
SWDEV-419688 - Do not run GWS init kernel for targets > gfx12 and MI300.
...
Change-Id: I8e7441268978be71ab8a5a33e7f8bcf69660e500
(cherry picked from commit 36d37ef614909c0f215512aac0c133408d787080)
2023-10-05 14:57:56 -04:00
e63c280d4d
SWDEV-422207 - Capture AQL Packets for graph Kernel nodes during graph Inst. And enqueue AQL packet during launch
...
Change-Id: I1e5f7f9e2a70bd500d190193cb6ba0867f5a63e7
2023-10-05 00:34:29 -04:00
d18ae85833
SWDEV-415914 - Use settings from device's setting object.
...
Change-Id: I4b0cbab361e9e2d5b8164066ec37de1fa212e783
2023-09-14 19:16:50 +00:00
42c9cd0320
SWDEV-301667 - Balance AQL packet scopes
...
- Acquire and Release scopes for AQL need to be balanced. These were
missing on the AMD_OPT_FLUSH=0 path.
Change-Id: Ibf4132eb96800f155d7b664359c790d68a353e60
2023-09-06 13:26:39 -04:00
1ec0ba3537
SWDEV-301667 - Use large signal pool
...
Use large signal pool if profiler is connected or profiling forced
enabled. This is needed to mitigate signal creation overhead when
profiling as signals are attached to every packet and deeper batch may
show overhead of signal allocation.
Change-Id: I8034b8a20b55328b87d593bf044f59672f9653e8
2023-08-24 19:17:05 -04:00
077311153a
SWDEV-407533 - [ABI Break]Purge unused env vars
...
Change-Id: I627950e8ebb6299affc602754a20d442dbe42b14
2023-08-24 14:11:40 -04:00
6f5277c701
SWDEV-408473 - Add wait time of 10 us if the waiting signal copy was < 24K.
...
Change-Id: I438ec9eb07e5034042a4a9a5e6e51d74daba2c83
2023-08-23 10:46:33 -04:00
63af50b6e0
SWDEV-415160 - Added fix in kernel argument logging
...
Change-Id: Ic54d51264602c7bf0c91b5273f9519da3fbb1062
2023-08-03 21:39:41 +01:00
0711320cb4
SWDEV-392732 - CP processes until doorbell ring
...
Change-Id: I3a0974435d3f95a40cea6ed326f744bcab5b4c0f
2023-07-25 13:10:31 -04:00
5447cf8872
SWDEV-301667 - Disable HostBlit copy for HIP
...
Change-Id: I46333ff42e8c1d402ece97e3ead7b539a27c3f82
2023-07-17 17:49:11 -04:00
b0e6f99ad7
SWDEV-392732 - Initial commit for graph doorbell optimization(AQL Buffering)
...
Change-Id: I451725006c54c249dc530c55d2af2a31594bf49b
2023-07-16 07:56:00 -04:00
bceec9d9a2
SWDEV-301667 - Enable SDMA profiling only when queue profiling
...
Change-Id: I9a62a2fbaefd54638c35f4b8fce2e9ea2da7af19
2023-06-26 16:10:34 -04:00
0330801846
SWDEV-406750 - Fix to avoid a warning when compiling HIP runtime
...
Change-Id: If1d848d334bc0121a96655781fc60b0321213398
2023-06-20 12:16:16 -04:00
5436d362b1
SWDEV-301667 - Add a flag for gpuvm kernargs
...
HIP_FORCE_DEV_KERNARG=1 will create a device allocation for kernel arg
segment. Flag is 0 by default.
Change-Id: Iaaf5a149f3be8596568878d5d272268baf067c60
2023-05-22 11:23:48 -04:00
06df9e2efd
SWDEV-301667 - Kernelarg gpuvm
...
Add aligned, nontemporal `memcpy` for kernarg.
Change-Id: I5d8ac76904feaf793b45ec2ea5fbd1069be20068
2023-05-22 11:21:14 -04:00
afa28cdf44
SWDEV-393362 - Move setData() from profilingEnd() to profilingBegin() to eliminate a possible data race
...
- Introduce a state variable to indicate if HwProfiling is enabled to
eliminate a possible data race of vector<> signals_.
Change-Id: Id504cc76d7fa9f7e6455587dd232b60ccbbb735b
2023-05-04 17:19:25 -04:00
c585aefdb7
SWDEV-376453 - Changes to fix crash seen with HIP and recent compiler change with -O0
...
Change-Id: I37de83b7449461ce2378741e97ff8fb8da736523
2023-05-02 17:46:38 -04:00
9775c38725
SWDEV-392427 - Invalidate Barrier Value AQL header
...
Change-Id: Id8e04ffe44da58641361468957d397af128443bb
2023-05-01 14:06:05 -04:00
20ca8b8116
SWDEV-384557 - Leverage SDMA engine status query
...
Change-Id: I5f386f2965de24a229ea43b6c4da82099692f91f
2023-04-05 07:50:53 +00:00
3e603d986a
SWDEV-364604 - Add ROCclr support for hipEventDisableSystemFence
...
Change-Id: I6127b432a8759359359a1890fda85bc401be6a56
2023-02-21 19:07:35 -05:00
695f169ed0
SWDEV-301667 - Refactor code
...
Change-Id: I8f20d994d88acb14f392eaa9d5cd14620667d7f6
2023-02-13 14:58:17 -08:00
67af16bdf3
SWDEV-381633 - Use blit manager on app's queue
...
Heap initialization used device queue, but it shoudl be used for
cooperative launches only. Heap initialization must use the same queue
as the current dispatch.
Change-Id: I856621bf82bbdeb1c2d0fbc4970e90d09af805cb
2023-02-09 09:43:13 -05:00
b23c759746
SWDEV-372790 - Copy AQL packet from runtime setup
...
Scheduler in device queue requires relaunching itself. Make sure
scheduler uses exactly the same AQL packet as the host launch.
Change-Id: I4eb03c4c91bf2408a6d4607731f081a2e2c2c8ae
2023-01-24 10:25:45 -05:00
53a10c9039
SWDEV-377991 - Remove liquidflash support
...
Change-Id: Iba6455e5c0210c3223a06fec332404cd9f489154
2023-01-20 09:57:06 -05:00
89e9f9db58
SWDEV-371123 - Fix seg fault when running hashcat app
...
Change-Id: I0c66d881b8dfae322f58a644860ecce4fb18debd
2023-01-19 15:50:03 -05:00
c8927cd84e
SWDEV-377991 - Remove Liquidflash extension
...
Initial check-in to untie dependencies with HIP and OCL repos
Change-Id: I363b63954c3f118f40a6ed893545d6a4ac44144c
2023-01-18 13:16:20 -05:00
77ceb7e9ad
SWDEV-371123 - Fix barrier value packet header
...
- Use correct header for vendor packet
- Pass one dependent signal when submitting a marker if there is one
Change-Id: I4efc70dd5204b559de26f899d0637f50421c8834
2023-01-12 04:06:03 -05:00
6651aab28d
SWDEV-371123 - Fix typo for log print
...
Change-Id: I18297a032d6fc9a030863eac7c525e86e9fdcd29
2023-01-05 17:46:22 -05:00
9076d9a518
SWDEV-366087 - Pass pitch and slice pitch to blit kernel from rect struct.
...
Change-Id: I1ffe54929db59a40e2a1ae19c125f8d8e81b07ec
2022-12-20 16:43:49 +00:00
72b45e2a1f
SWDEV-369581 - Convey copy API metadata to ROCclr
...
Change-Id: I569462d6d268700d419510255e201bf7d80d6714
2022-12-09 00:27:15 -05:00
4f64d89026
SWDEV-371123 - Use barrier value packet for event records
...
Change-Id: I5e5e5e89e0d96a2430b4682d168b76848fa5b94e
2022-12-07 17:57:36 -05:00
5d7f3f9f3c
SWDEV-305894 - Cooperative groups grid and multi grid sync support for gfx940+
...
Change-Id: I35d72f1cb50c3a96eee56a612b72d641852b145f
2022-12-05 16:30:30 -05:00
10ecf8f159
SWDEV-371123 - Refactor dispatchBarrierValuePacket
...
Change-Id: I846bd3f60dd8db125e2ca5475e9d25fdd567922b
2022-12-02 12:22:31 -05:00
61b6a94627
SWDEV-358915 - Wait for 10 micro secs, if the next engine is SDMA.
...
Change-Id: Ib3ed91510e62090f47d4f55be8de635caaee083d
2022-11-28 12:14:16 -05:00
feca11d5e3
SWDEV-301667 - Improve logging
...
Change-Id: Ifa6da876b85cb503967cf09aac6d477b10db8e63
2022-11-04 18:23:18 -04:00
3e465c5ff8
SWDEV-364052 - Print free memory on queue callback
...
Change-Id: I7d15b6f6277326d5f3e9b784a3443f18ec6ba58a
2022-11-04 13:19:33 -04:00
b043b4f5a2
SWDEV-351980 - Store the correlation ID in the dispatch packet
...
The profiler can retrieve this correlation ID to attribute waves to
specific dispatch locations.
Change-Id: I700e8a91219d612f6a2028c0dda0c92753f3526a
2022-11-02 23:40:10 -04:00
166ef6b622
SWDEV-293816 - Only dispatch Counter Aql packet when it is created successfully.
...
Change-Id: Ic8a7374d19012754c6de5d483ec5d07a56661d1b
2022-10-31 15:11:41 -04:00
cf1fd98074
SWDEV-301667 - Use queue base address for AQL log
...
Change-Id: Ie525991b341deeb93ab3cb2d77550e8e1d87ee20
2022-10-31 13:54:49 -04:00
b2665ccc4b
SWDEV-361730 - Fix to match the kernel arg allocation sizes between gpu visible mem and host/runtime mem
...
Change-Id: I6495aa57e97b46373c1d90cc230435780fdcf3ef
2022-10-20 14:20:37 -04:00
Saleel Kudchadker
Anusha GodavarthySurya
Ioannis Assiouras
kjayapra-amd
Jaydeep Patel
German
Sourabh Betigeri
Alex Voicu
Jatin Chaudhary
Laurent Morichetti
jatang