Commit Graph

8920 Commits

Author SHA1 Message Date
eshcherb 036f7d48a1 Update README.md 2018-11-13 18:51:56 -06:00
Evgeny b7467b9995 preloading local libmcwamp_hsa.so 2018-11-13 18:44:44 -06:00
Evgeny 48d69df3ee packaging 2018-11-13 17:09:58 -06:00
Oak Zeng acb80d7583 Create SDMA queue on specific engine
Change-Id: Iece03795510d66b03324174203faa0ac9eb4fb7d
Signed-off-by: Oak Zeng <ozeng@amd.com>
2018-11-13 14:52:57 -05:00
Oak Zeng 8d65e72045 Move m_Type to a local variable
BaseQueue class has a member function GetQueueType so m_Type
is duplicated.  m_Type is only used in one function. Move it to
a local variable.

Change-Id: Ice144cf723178dd628cb49261c23d10605f9ee7d
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
2018-11-13 14:52:17 -05:00
Oak Zeng 58b95e0a9d Create SDMA queue on specific engine
Change-Id: Id651ececda55b81b45e991bd8e6616674be48d8e
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
2018-11-13 14:52:17 -05:00
Oak Zeng 5173e71810 Add more SDMA queue type
Those new types are used to create SDMA queue on specific engine

Change-Id: I91c3bcc14fef7404cf42b256a18651432e171091
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
2018-11-13 14:52:01 -05:00
Oak Zeng 055f7c9c2c Use latest kfd_ioctl.h file
Change-Id: Icd7da4a305581c6857e17d59fbd0c3bd5101df3b
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
2018-11-13 14:51:46 -05:00
Sean Keely 4e8597681b Cache KFD Events used by user allocated InterruptSignals.
Change-Id: I7f102f880fea9c78febe28cd262f93ee77f03184
2018-11-12 22:37:42 -06:00
Sean Keely 8323b2e1d7 Add pooling for Signal ABI blocks (SharedSignal).
Makes better use of memory and greatly reduces mmap count.

Change-Id: Ib444cd1ccd144986adbcc7cec297a966e2c08bc7
2018-11-12 22:37:28 -06:00
Chris Freehill 861c2c2e33 Add rsmi_dev_busy_percent_get()
Also: correct some comments, ifdef out unused code
2018-11-12 17:25:14 -06:00
Felix Kuehling 5e4e19d47b libhsakmt: Distinguish EPERM and EACCES
EPERM means "operation not permitted" and is returned when CGroup
access checks fail. EACCES means "permission denied" and is returned
when the device file permission bits or access control list don't
allow access.

EPERM can fail silently, since we assume the administrator disabled
a device on purpose in the CGroup. EACCESS should produce an error
message and an info message to check the device file permissions.

Change-Id: Iee4c5584c5fdc4e113c3d760dede6661097b4341
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
2018-11-12 17:06:18 -05:00
Evgeny 722fa91624 pre/post package scripts install dir name fix and CPACK_PACKAGING_INSTALL_PREFIX
Change-Id: I415875572829dadc9996dd7561418a77ec3fe1bb
2018-11-12 12:39:03 -06:00
eshcherb 3ddb17864f Update README.md 2018-11-12 11:26:10 -06:00
Evgeny a52d7b36a9 pre/post package scripts install dir name fix 2018-11-12 10:16:38 -06:00
eshcherb 48db2cddc3 Update README.md 2018-11-12 00:19:31 -06:00
Evgeny 270613e3a6 activity id fix 2018-11-11 23:33:36 -06:00
eshcherb 1bceac618b Update README.md 2018-11-11 18:40:07 -06:00
Evgeny 44986fcbda cmake packaging fix 2018-11-11 00:53:10 -06:00
Evgeny 6b3719051f package generators deb/rpm/tgz 2018-11-10 21:33:34 -06:00
Evgeny baf2f25bac packages deb/rpm/tgz
Change-Id: I1de06a8ad19a5c44baf45a76fa34db570899c72d
2018-11-10 21:28:51 -06:00
Evgeny d4f03f0244 removing HCC_PATH 2018-11-09 23:56:28 -06:00
Sean Keely 936ecd1885 Remove legacy SVM region concept.
Also rename blit_agent to region_gpu and add comments to clarify
its role in deprecated region API support rather than to do blits.

Change-Id: I80b1043db2e1c5d40a58fc801eef70a688ea9169
2018-11-09 06:27:53 -06:00
Sean Keely dda9c17b45 Move VM fault handler init to after all devices are registered.
During registration we must not call any function that depends on registered
data as the lists are not yet complete.  This includes signal allocation since
allocating shared GPU mapped memory depends on the list of GPUs.

Change-Id: I94d59e847802c546c2a5a0d9f55fe5ac3fd1d878
2018-11-09 03:10:08 -06:00
Evgeny 06e268fbc1 hip_prof_(api/str).h moving under hcc_detail dir 2018-11-08 05:01:01 -06:00
Evgeny 1d55085060 cosmetic changes
Change-Id: Ie9c730305c8a882e250f28cc60d640c70d899572
2018-11-08 03:54:09 -06:00
Mike Li 3afce42b57 Changed scripts to include running kfdtest in docker container
Change-Id: I822ff4869610df6abad846542d7c290b7a5aae79
2018-11-07 16:09:12 -05:00
Evgeny 4bdb8407f8 tracker handlers ordering
Change-Id: I998ff13380eaa3d920a0037951acef596d850dfd
2018-11-06 00:53:09 -06:00
Chris Freehill 59a952666f Add rsmi_dev_pci_id_get() to return BDFID for given device
Also:
* add some exception handling;
* chop newline character off of device name returned from
rsmi_dev_id_get()
2018-11-05 11:22:12 -06:00
Evgeny dd19d85644 API normalizing 2018-11-05 02:31:37 -06:00
Evgeny 812c17ed77 rpl_run data extracting error message; comments for intermediate metrics
Change-Id: I4d9f97ed475d3ec328aafae40f1bc80b6bf69bcc
2018-11-01 13:38:01 -05:00
Gang Ba c54c1dbdcb Add code to support packet capture and replay in the Thunk
This feature only support dgpu for now.

Change-Id: Ic766ec06892c955dd605ecc335a776335edc0df2
Signed-off-by: Gang Ba <gaba@amd.com>
2018-10-31 16:53:46 -04:00
Evgeny f977ac2fbf rpl_run w/o input file; queue create callback; test for n gpus and n threads
Change-Id: I37157c49cf6454de591cae97b5cc43287ea95956
2018-10-30 14:19:45 -05:00
Harish Kasiviswanathan c1994e28f0 libhsakmt: Support device controller cgroup
Device whiltelist controller cgroup allows to track and enforce open and
mknod restrictions on device files. Tasks should works with
/dev/dri/renderN devices that are whitelisted for its cgroup. If a
certain node is not whitelisted it is not an error condition.

Change-Id: I0b997423ccdc00aee98df5b6f04ed6794549604e
Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
2018-10-30 11:31:53 -04:00
Chris Freehill 62ba2f578e Use sysfs file to get average power instead of debugfs 2018-10-29 17:59:24 -05:00
Jenkins 460f2e6b1c Merge 'master' into 'amd-master'
Change-Id: I3ef383f30735dea9eaa62e114cc39aa35f7cc741
2018-10-26 04:09:48 -05:00
Chris Freehill 0c84f5117e Merge "Merge 'master' into 'amd-master'" into amd-master 2018-10-25 15:42:28 -04:00
Chris Freehill 767fa53d8c Add support for new performance levels
Also added tests for new performance levels and clean up some
formatting/style issues.
2018-10-25 14:13:55 -05:00
Kent Russell 10edccb912 Specify requirement of NUMA libs for Thunk
Add the numa libs to the thunk specs for DEB/RPM, so we can remove the
manual installation requirement


Change-Id: I5aadcf581b64e9a20aee9c1e1204af4715d1e990
2018-10-25 07:37:07 -04:00
Jenkins 6d30a7f126 Merge 'master' into 'amd-master'
Change-Id: Ia3824075144415287fbe67fff68a52e12f10f5c5
2018-10-25 04:09:39 -05:00
Jenkins 41eb15e424 Merge 'master' into 'amd-master'
Change-Id: Idf74c08d56275eea693955c9380a5330d54def67
2018-10-24 17:39:24 -05:00
Chris Freehill b9c6a5a443 Add gtest based testing 2018-10-24 11:57:03 -05:00
Philip Cox 105edd4bb4 Fix Debug Thunk spec mismatch
Move debug trap support capabilities to their own
structure to fix thunk spec vs header mismatch.



Change-Id: I6694601bfa36097502c8ab932e082d7a4645d5b2
Signed-off-by: Philip Cox <Philip.Cox@amd.com>
2018-10-24 11:32:12 -04:00
Sean Keely 9ec37b5103 Ensure runtime cleanup when hsa_init ref count reaches 0.
Delete the runtime object when the last hsa_shut_down occurs.

Change-Id: I2005d52d06702eaef166714fd5e471cc277924db
2018-10-22 19:32:00 -05:00
Chris Freehill b2f1ee1764 Comment corrections, debug output and assert added 2018-10-22 17:26:38 -05:00
Evgeny d788a53972 aqlprofile extension version check
Change-Id: If824764f199eca15a0341cdf6177d8d6353e29f3
2018-10-22 15:36:57 -04:00
Sean Keely 757502ccd6 Report internal queue creation to tools.
Debug agent requires handles to internal queues for single step debugging.
Added tools only API hsa_amd_runtime_queue_create_register for reporting.

hsa_amd_runtime_queue_create_register sets a callback which is invoked
when internal queues are created.

Change-Id: Ia5190ae724fadba686c15f25b2cd085350eeff0e
2018-10-20 23:12:27 -04:00
Sean Keely 5975c465ad Fully initialize GPU agents before loading tools.
Required for debug agent requires copy API and trap handler to be initalized
prior to loading.  Existing tools do not make use of internal queue or scratch
memory intercept which is what PostToolsInit allows.

PostToolsInit() will be removed in a following cleanup change.

Change-Id: If43377843808e3eff0defd9204910a67a852902f
2018-10-20 23:12:14 -04:00
Sean Keely 6852282a07 Refactor of Runtime::CopyMemory()
Change-Id: I32a7cb24d00660ff4471d121ef7b3c2eec8fced2
2018-10-20 14:38:50 -04:00
xinhui pan 7a13bb4d66 kfdtest: blacklist KFDQMTest.SdmaEventInterrupt
On gfx900+, the test sometimes timeout due to cp fw bug.
Blacklist it until we address the root cause and have a fix.

Change-Id: Iff600a6f6dbd86c56e034f530484205520bced32
Signed-off-by: xinhui pan <xinhui.pan@amd.com>
2018-10-19 15:29:54 -04:00