Commit Graph

8920 Commits

Author SHA1 Message Date
Evgeny 68c9f32d6e fix output merging
Change-Id: I7c51e85f8a8e6e5bae67abf74716a658cc533a6a
2020-06-15 01:13:19 -05:00
Chris Freehill 8e6f7c798d Don't automatically overwrite manual .pdf file
Automatically updating the manual pdf file causes a local
git change. This messes up "repo sync" calls because of the
local change. Instead, just write an un-tracked file that can
be used to update the tracked version of the manual .pdf.

Change-Id: Icd7edc244df60728ec169c5aa1cf8b322ca4143b
2020-06-12 14:19:15 -05:00
Tony 047b80ab41 Do not add offset/size to code object URI if loading from complete file
Change-Id: I7cb1e803e0bcdbc24935273af431fbcc88404bfd
2020-06-12 05:10:35 -04:00
Rachida Kebichi 807f17d6ce support for hip multi kernels
Change-Id: I8cc42eb53f075ea4f27dc1287e7ab3fb77916027
2020-06-11 14:59:22 -05:00
Sean Keely f0b0555996 Add optional support for the cmake per user package registry.
Default is OFF to conform to latest cmake standard (3.15) and
because this feature can cause some confusion for unaware developers.

Change-Id: I6fdbd5ab76c90b73ac6451a383334a83cabef438
Signed-off-by: Sean Keely <Sean.Keely@amd.com>
2020-06-11 14:25:57 -05:00
Sean Keely 06add334b9 Update to conform with modern cmake best practices.
Immediate benefit is that this enables cmake native find_package
support including transitive passthrough of link dependencies,
options, and include paths.

It does require a bump to cmake 3.6 to enable proper package file
names.

Change-Id: I3f4e2f2db745c04a7f686111b080098a0d92fbf5
Signed-off-by: Sean Keely <Sean.Keely@amd.com>
2020-06-11 14:25:41 -05:00
Rachida Kebichi 64ee8f29ef support for hip multi kernels
Change-Id: Idc8945f924ceb35081e153921181b96ee5f653b6
2020-06-11 15:13:33 -04:00
Evgeny b033453537 concurrent sqtt support
Change-Id: I0e1515d4b0aab30b0a9b673120eee28d96f64d9e
2020-06-10 22:45:30 -04:00
Chris Freehill f946ea37ef Update XGMI perf counter test to show utilization
Also:
* When destroying a counter, make sure to stop the counter first
* In the test, do not stop (disable) the counter before
  reading it.
* Clean up some whitespace in other tests
* Re-add manual pdf file

Change-Id: I0786ef3a994ca568299c77e44f092af8943ac33d
2020-06-10 12:49:49 -04:00
Evgeny 5a0fd11cf8 run_ci fix
Change-Id: Ibf07cb92c55790be723dc2e97d09f3232983747e
2020-06-10 10:01:35 -05:00
Evgeny f2d3fd943d packaging fix1
Change-Id: Ifc8941b2a720628d76e0697debe6a861fa349b2b
2020-06-09 19:31:37 -05:00
Evgeny c4e3d8de32 JSON GPU activity per stream fix
Change-Id: I3fca861d0b36d26d2800cd39f12cd4485a2b4db9
2020-06-05 10:16:57 -05:00
Evgeny Shcherbakov a203dc041d Merge "CI run script version" into amd-master 2020-06-04 20:31:41 -04:00
Evgeny d2135ad2c8 CI run script version
Change-Id: I0e806854eb7db12656f7f1770642f20247fa6734
2020-06-04 17:31:18 -05:00
Evgeny Shcherbakov 6a96c90310 Merge "adding proc-id to activity events" into amd-master 2020-06-04 15:03:34 -04:00
Kent Russell 8cf44548c0 Make an empty unique_id file non-fatal
This isn't supported on all models, so just comment out on failure
instead of fully failing

Change-Id: Id36d5df7c87abbb41f7b6be43abfea82004703a6
2020-06-04 10:31:53 -04:00
Tony a74660c69a Initialize HSA Queue object before creating KFD queue
Initialize all the fields in the HSA queue object to known values
before calling the thunk to create the KFD queue. This ensures that
when the debugger detects that a KFD queue is created it can access
the values it requires. The values it requires include the apperture
addresses, queue scratch memory base, and the HSA queue kind.

Change-Id: Ic985755b0402c6794d5987e60aff50d223f09eb9
2020-06-04 04:56:24 -04:00
Konstantin Zhuravlyov 5f614c31f5 Correct loader URI reporting
- Check address is in the range of the mapped file.
  - Correct calculation of offset within the file.

Change-Id: I848a3ead4422698c2ef1c140bc8ae5e000a717f7
2020-06-03 13:03:07 -04:00
Evgeny 35e6ac1d0f adding proc-id to activity events
Change-Id: I3a4673c8e5d8b115a213d00058e4235a4e68510e
2020-06-03 03:29:24 -05:00
Evgeny 84ed8c4b63 adding proc-id to activity events
Change-Id: I83b587651819a7cb08ffdd5f85bfe82927521720
2020-06-03 03:26:36 -05:00
Evgeny e9a9b5d3ef rocminfo check fix
Change-Id: I8d50335060a16954df37be8f4d74928d565b1078
2020-06-03 01:00:48 -05:00
Evgeny Shcherbakov 03a2a761b3 Merge "sqlight generic fld change api" into amd-master 2020-06-02 22:32:59 -04:00
Laurent Morichetti da6d892058 Handle address watch exceptions
Set ttmp11[8] and send a signal for the debugger when the handler
is entered because of an address watch exception.

Change-Id: Icc83a79027bb7ca1e50e19e2f00464cb9ca862f3
2020-06-02 15:25:15 -04:00
Evgeny 6f0e4dbc61 sqlight generic fld change api
Change-Id: I1ff21da94bd61a3aecf2a54f0212cd3b88b5776e
2020-06-02 11:00:19 -04:00
Evgeny 07479cf881 adding tracer flush/unload stage unloading
Change-Id: Id84774b2ce20b261a9a84d2ad802b76093a748d4
2020-06-02 05:10:02 -05:00
Evgeny 0efdd937a2 minor change
Change-Id: I365e3c671c829c6b205587f5770f70cef5bbddfb
2020-06-02 06:07:08 -04:00
Evgeny 59a4ae9abb list-derived/basic fixed
Change-Id: I2038eac424b1485c304ffecf35ba2a0965de5f3b
2020-06-01 22:04:37 -04:00
Evgeny 1c11cd087b ocl samples disabling
Change-Id: I8012e226f96246ed70974048a87e67da9d01753c
2020-06-01 18:07:52 -05:00
Evgeny 8d0801de75 run test environment
Change-Id: If4b780357c807f88d09d7728d13d3c616b424474
2020-06-01 12:54:35 -05:00
Konstantin Zhuravlyov 9eb735ec24 Add support for code object URI to ROCr
Adds the following:
	- New factory method to create a code object reader from
          file with offset and size.
	- A pair of queries on a loaded code object to get the URI name/length.
	- A bump to the AMD vendor loader extension API and its associated table.

Change-Id: I17c83e9c2447d29a43c438459395365f786a3611
2020-06-01 11:07:50 -04:00
Evgeny Shcherbakov ce5b5ed9bc Merge "setting code-obj tracking by default" into amd-master 2020-05-30 03:34:46 -04:00
Evgeny Shcherbakov 6b58ea8161 Merge "fixin build: check script residing in build area" into amd-master 2020-05-30 03:34:08 -04:00
Evgeny 1e7650074a fixin build: check script residing in build area
Change-Id: I1a3d794b24c7506d5e24c341c61fe45b31238002
2020-05-29 16:09:13 -05:00
Mukul Joshi 633c852f5d Print VRAM usage in rsmitst
Print VRAM usage information in TestProcInfoRead.
Also, fix output formatting when running TestProcInfoRead.

Change-Id: I9efed808458ef4645145610f6f564f0f2baadea2
2020-05-29 15:48:06 -04:00
Evgeny 511f659a8e hipLaunchKernel kernel name fix
Change-Id: I59c63ec34f77fe39ed1edc8b6395db3871004e40
2020-05-29 13:16:14 -04:00
Gang Ba d675d1cce1 libhsakmt: add Streaming Performance Monitors APIs
Signed-off-by: Gang Ba <gaba@amd.com>
Change-Id: I5c23a8dacf9bc50c740908aabe391432f2c7112e
Signed-off-by: Gang Ba <gaba@amd.com>
2020-05-29 09:34:31 -04:00
Sean Keely 4e0bf29704 Separate memory map info printing from vmfault handler.
Allows the print utility to be called from GDB.

Change-Id: Ia27ff9ba23ad6c0117ba1278693d4c8b0473bdc1
2020-05-28 23:34:49 -04:00
Sean Keely 32bb10086d Select a non-empty pool for image kernarg.
Zero size pools have no numa bindings.  Selecting a pool with numa
bindings should prevent thrashing due to numa balancing daemon.

Change-Id: Ib0082cb9af66e24e07a2adbb83c1045145d51403
2020-05-28 23:34:42 -04:00
Sean Keely 40d1931209 Relocate initial system clock setup.
This has nothing to do with registering agents.
Moved to Runtime::Load.

Change-Id: I0f84c9d8f5a68d458717111113f02af56c92f4f6
2020-05-28 23:34:30 -04:00
Sean Keely 29b660c91e Correct IPC fragment reuse.
Memory from the suballocator may be exported via IPC.  If this
happens then the allocating process should not reuse that memory
since it would still be connected to the remote process.  IPC exported
memory must be released back to the driver.

Change-Id: I2ab0c814f63191f753fc3640cc4140ee144bf07f
2020-05-28 21:46:02 -04:00
Sean Keely 09ebc21d13 Remove PointerInfo fragment branch.
All types which could be generated from a fragment need to take this branch.
Taking the branch is correct for all types, it was a performance optimization
only and was missing IPC.  Branch removal simplifies updates for any future
fragment use and will allow CQE to report any performance issues that might
require bringing the branch back.

Change-Id: I8041788c422e880b764e144eb1877f5126ba76f3
2020-05-28 21:46:02 -04:00
Sean Keely 397608e2c0 Fix pointer info block base address fetch.
Thunk may report nullptr for host base if the host does not have
access.  Use agent base in this case.

Change-Id: I44883d35a3fff0941b1e3037d16b059591a6c511
2020-05-28 21:46:02 -04:00
Evgeny 201fc86f5c workaround for hipLaunchKernel not having kernel name
Change-Id: I7d165a7d897cced7b6045f084d2112a0e9e33e69
2020-05-28 16:58:33 -05:00
Joseph Greathouse 3fe95c696a Add ability to get ASIC revision from hsa_agent_get_info()
Change-Id: Ie30c64f9e02f9e704c9a4c4145deb9580429fdf6
2020-05-28 14:25:13 -05:00
Evgeny 30db99e758 setting code-obj tracking by default
Change-Id: Id6a97a7dc77faa3b7eb0e2b81b75c13ca7fc5818
2020-05-28 03:43:12 -05:00
Evgeny 157bceebdf testing fix
Change-Id: I4950c3ddeb79947eb311c4f008ad91b7311bb9b3
2020-05-28 02:17:01 -05:00
Chris Freehill 42c10633b6 Check only the minor ioctl version for event support
Change-Id: I70ddab4298a62178b2509a0365ee4cd6937302c1
2020-05-27 09:27:01 -04:00
Evgeny 9d4a58f6da disabling kernel name printing for kernel launches by func pointer and ostream operator template
Change-Id: Icbcedafb1812a1be03837f6cb47283eb0d382ad2
2020-05-27 05:23:29 -04:00
Evgeny 4debecbc3c removing /opt/rocm hardcoding
Change-Id: Ifb041a76a0c6fa943f72a9228197f371ad384bb7
2020-05-26 21:56:07 -05:00
Mukul Joshi e30ebbc787 Add support to retrieve process VRAM usage information.
Change-Id: I60843a99207a658022a26aa346b79f91863833cf
2020-05-26 15:19:24 -04:00