Commit Graph

8920 Commits

Author SHA1 Message Date
Evgeny 21d6c00bb6 more features 2018-05-25 22:39:22 -05:00
Evgeny 5151f6ce04 betta version 2018-05-25 01:56:25 -05:00
Evgeny b8fc3818de adding dispatch_packet field to rocprofiler_callback_data_t
Change-Id: I3cbe256130fa23dd24b094f27286d566ed339230
2018-05-22 13:17:03 -05:00
Qingchuan Shi 3a46556dcc Add debug trap rocrtst.
Change-Id: I73682d7a2ad51eed9988075e012478a1afc76c7c
2018-05-22 13:31:45 -04:00
Yong Zhao ec440fb428 Stop allocating eop buffer for SDMA queues
Change-Id: I9a4eaee05588292a797eb424503dd7b793c1408c
Signed-off-by: Yong Zhao <yong.zhao@amd.com>
2018-05-16 15:30:23 -04:00
Yong Zhao 43f119bcbc Improve the code readablity
The main point is to move update_ctx_save_restore_size() out of if()
condition.

Change-Id: I58a1a4f3edca2d1c510fdd0e31e59b5c41e92a14
Signed-off-by: Yong Zhao <yong.zhao@amd.com>
2018-05-16 14:55:55 -04:00
Evgeny 2d4d704f56 sync and clean up
Change-Id: Ia3b2d76f9016454b9e80ca0891ee02e0c2fd1a13
2018-05-16 13:24:04 -05:00
Evgeny cca3ca68c1 aqlprofile/rocprofiler util sync
Change-Id: If029a5aa4eff9ad3e01dc83b41426c079f7f03a7
2018-05-16 13:04:07 -05:00
Evgeny 64ef6bfd07 fixing backward compatobility, group_t structure
Change-Id: I0b188a055638ed542e1161ff4650eedb7ab1bd0f
2018-05-14 11:09:33 -05:00
rohit pathania 47af1d673e Memory Concurrent tests for pool Memory allocate, Memory free and get pool info
Change-Id: I6a1343348e400fe466e041d651adaa67be561a21
2018-05-14 01:30:54 -04:00
Evgeny ee77feb186 revertingROCP_HSA_INTERCEPT env var
Change-Id: I1bf41f5773d194ac409056271a15b61366cc6292
2018-05-10 17:37:54 -05:00
Evgeny 0a934cff68 Initial commit 2018-05-10 13:19:10 -05:00
Jay Cornwall 536823482b Handle llvm.trap only in gfx9 trap handler
llvm.debugtrap and other trap IDs are reserved and should not place
the queue into an error state.

Change-Id: I98193a35ac7da94c4a42ee75d87754ee552ebea0
2018-05-04 13:15:23 -05:00
Felix Kuehling 2915d521a1 Remove redundant dev package build
No need to build the package in the build-dev target. This is taken
care of by package-dev. Removing the redundant packaging command
allows install-dev to work without building a package unnecessarily.

Also moved the rm command into the package-dev target.

Change-Id: I044871be03ebc5673146b44e4291b48b112f4440
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
2018-05-04 14:04:19 -04:00
Evgeny 0e0be791ec Tool load failure report changing to unconditional print bcos it's already is controlled with the env var
Change-Id: I91b400ba94575a32005e825e6b41bda05c55b357
2018-05-03 22:31:17 -05:00
Evgeny 8bed188f6a adding allocate cmd memory
Change-Id: I5fcc18673eba92e4ff84eb97ddfb16e6797eaf0d
2018-05-03 18:50:29 -05:00
Evgeny 71044909ce test structure refactoring, unifying with aqlprofile
Change-Id: Icad642413c7edb27885effc9da3524cf3899f8de
2018-05-03 16:47:18 -05:00
Evgeny 1769fea212 test migrated to new kernel OCL ABI
Change-Id: I2cc281316636f7b80dec2c2a78aade39a8ed168c
2018-05-03 14:42:46 -05:00
Oak Zeng dc1bbccc39 Use svm aperture for device memory allocate for gfx902 and after APU
Change-Id: Ib1d822adde30138a016e010bf581220465a087b9
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
2018-05-03 12:03:22 -04:00
Sean Keely a20cceb425 Relax large scratch cutoff.
Optimize for few queues rather than max.

Change-Id: I6531427319d3b2109b70d98fdb51daee7ffe4403
2018-04-30 07:25:22 -04:00
Sean Keely 5f25619bb7 Enable large scratch on GFX8.
Ensure system release fence is set on GFX8 large scratch using packets.

Change-Id: I13cfdcd35969482ea6e95e0b352f5cb3a0454b86
2018-04-30 07:24:53 -04:00
Evgeny 9dec361cd4 SQTT local allocation
Change-Id: Ie4a150ad0dc141226f6f1c571916c5a526dd723c
2018-04-29 03:24:46 -05:00
Evgeny 0fcd2fa56e aqlprofile: get version API
Change-Id: I3a85c088bfff3f54d8829e17cdafd7dfcdfb0c1d
2018-04-27 23:11:16 -04:00
Evgeny c9c0ecc976 memory allocation refactoring
Change-Id: Ic63b4f5ea44f2dc5e009e3e58652a661e957b7d6
2018-04-27 20:00:20 -05:00
Chris Freehill 11e13704ea Added rocm-smi-lib calls with updated interface
Change-Id: I62b59dca5135ec012f11b249c78b04e5e8e2dd9a
2018-04-27 18:36:54 -04:00
Evgeny b37027e347 aqlprofile: read API
Change-Id: I896b1fbf1c19608197ac0a99b9d467d8c1bee775
2018-04-27 11:00:08 -05:00
Sean Keely 7cd6e366ed Workaround SDMA poll packet preemption.
Use async. signal handler to satisfy dependencies for SDMA blits.

Change-Id: Ifa8d3ee6810509f400a568ca2387ac6ab3ab7c36
2018-04-26 02:00:33 -05:00
Evgeny 0d286e5efa the timestamp to ns conversion correction
Change-Id: I4928519dcef5850a9e43670d0e916e52ebf8cb0b
2018-04-25 16:57:18 -05:00
Evgeny 729ce983cb timestamp fix, converting to ns from system timestamp 2018-04-24 19:10:46 -05:00
srinivas Charupally 8c8cd2dbd0 Adding concurrent init and init shutdown tests
Change-Id: Ifdbda16ae6c93a86373557f26eb414e40775d343
2018-04-24 11:56:50 +05:30
rohit pathania b1b036acb8 Memory Allocation Negative tests
Change-Id: Icdd355f2351968dd76a3bc466636e223573cfb16
2018-04-23 01:49:43 -04:00
Qingchuan Shi 49d2175c74 debug suport for queue error.
1/ Revised debug event handler to handle different events.
2/ Added queue error handler using the callback in queue create, which will print out wave info when queue in error state.
3/ Preempt queue instead of destory queue when queue error state.

Change-Id: Ib727d208de9caf1c72c76d42268483b24aaebde8
2018-04-20 14:25:16 -04:00
Evgeny d04f7095f4 HSA memory alloc/copy/free refactoring 2018-04-20 11:15:26 -05:00
Sean Keely b66764e4c6 Disable large scratch on GFX8.
Temporary pending firmware fix.

Change-Id: Id1b1ecef421bc97327fd0d2e6225549a6e81dba0
2018-04-19 20:26:12 -05:00
Evgeny b2bf09b9e0 adding SQTT SE_MASK parameter
Change-Id: Ifbccaeb735a68ab77c70fb6fd975520d037d29d4
2018-04-16 17:25:40 -05:00
Evgeny 5a6f47c475 aqlprofile API: adding SQTT SE_MASK parameter
Change-Id: I0149692c2249c6d84ca710ce64e7346784ae593f
2018-04-16 16:39:42 -05:00
Evgeny 62531b88ed rc file options: 'basenames', 'timestamp', 'ctx-limit', 'heartbeat', 'sqtt-size' 2018-04-11 13:26:58 -05:00
Evgeny c096c31a2c include directory fix
Change-Id: Ie260f607a1c5f5e3e40375a8adeffefd9fb28ec4
2018-04-11 00:07:54 -05:00
Hari Thangirala 3e0cd85d69 Allow HSA_ENABLE_SDMA to override runtime defaults.
Change-Id: I2305304228010157bfb589c365f4a998577231cd
2018-04-10 12:56:48 -04:00
Konstantin Zhuravlyov 7ef70f7eaa Bring naming on par with the spec (hsa-runtime)
Change-Id: Ie1903c90a195cf95b186eb5552131a20af408adf
2018-04-10 09:15:02 -04:00
Evgeny cf11e49e2a rocprofiler-dev package
Change-Id: I80c4f6fd0619f4985516be66dedd2945abcdf9af
2018-04-09 09:05:56 -05:00
Chris Freehill 0d9e71a63a Revert "Re-enable rocm-smi with new c api"
This reverts commit 2e81d33395.

Change-Id: I9866610597a6de97a3c06ef9646f0afc85f149f4
2018-04-07 19:59:13 -04:00
Chris Freehill 2e81d33395 Re-enable rocm-smi with new c api
Change-Id: Idf393f31522bac8ac0c3c03a930ef66d97ce5fa2
2018-04-07 18:55:30 -04:00
Ramesh Errabolu f25d59cab2 Compute size of command buffer based on support for HDP Flush
Change-Id: I4987a262c191a91cd845fe18002c314a95a9ed8c
2018-04-07 13:36:09 -04:00
Sean Keely 7caf9633f6 Support large scratch allocations and reclaim.
Also improve small_heap used for scratch region allocation.

Change-Id: Ib7311b663b38968d88ebc355b81e12c0863dc541
2018-04-05 21:51:56 -04:00
Evgeny f5d6d78a21 make clean fix 2018-04-05 18:45:20 -05:00
Evgeny e4384c2c51 loading of config rpl_rc.xml file; range format [N], [N:], [N:M] 2018-04-05 18:03:05 -05:00
Evgeny ee5bc95019 SQTT memory checks 2018-04-04 12:14:29 -05:00
Jay Cornwall df964343a3 Support new first-level trap handler ABI
- Ignore exceptions passed to the second-level handler
- Restore SQ_WAVE_IB_STS and SQ_WAVE_STATUS before exiting trap

Change-Id: I872c111c030d94eae644ae073df3c2e508f42f45
2018-04-04 11:01:14 -05:00
Sean Keely b6f0248f53 Respect new memory model requirements at queue destroy.
Spec requires GPU release fences and CPU acquire fences at queue destroy.
Also update the recognized status codes.

Change-Id: If9166f5149f65417c7057ff7c0f69f6ac094d6ab
2018-04-04 08:13:00 -04:00