bdb5edad34
Fix memory leak in exception path.
...
Change-Id: Iad5f035cd1909be4a8f1a1f5dd7ca5abec0694b4
[ROCm/ROCR-Runtime commit: 737966eb25
2017-10-21 11:08:00 -04:00
a4d753615b
Changes to support surfacing of link weight as part of link info
...
Change-Id: I1c0705a9374af1245f0419c51beded0d7ee10639
[ROCm/ROCR-Runtime commit: dccbc9f2af
2017-10-20 12:09:31 -04:00
26ebd1727c
Use Major/Minor/Step device numbers to differentiate gfx devices
...
Change-Id: I0901871971a5b33018917ada6c0e69ac7aa91944
[ROCm/ROCR-Runtime commit: a7cbe78366
2017-10-13 16:18:24 -04:00
93e58b9100
Capture more memory allocation types with the 2MB allocator.
...
TensorFlow was running out of VRAM due to padding up allocations
from legacy memory APIs. These allocations have been added to
the fragment allocator to improve VRAM utilization.
Change-Id: Ic680fff576a0434b3b17a4c91746da44e09957fa
[ROCm/ROCR-Runtime commit: 4f299a9909
2017-10-12 23:22:10 -04:00
b07cbdc10e
aqlprofile API: enabling privilege memory related counters
...
Change-Id: I28a24ad1a3ce78c5d8a6319635ae1ffd392ab690
[ROCm/ROCR-Runtime commit: fd99e909ff
2017-10-09 17:34:54 -05:00
06c7997137
Release cached memory blocks when memory allocation fails and retry.
...
Change-Id: I6d96e136e147d8ffe9ff7baec26b4b5a678b739f
[ROCm/ROCR-Runtime commit: 9ba83d83f7
2017-10-03 20:12:05 -04:00
21a65a1233
Support hsa_amd_agents_allow_access on page fragments.
...
Since access may only be manipulated on whole pages, suballocator fragments must cooperate to set the page's access.
Since the KFD does not migrate memory on access changes this implementation makes agent access sticky across the requests in a fragmented page.
Change-Id: I88479ed45fb40e9782b704526a7b8ffb22e7bd76
[ROCm/ROCR-Runtime commit: e9a6f2c3e6
2017-09-27 19:04:04 -05:00
64b516415c
removing graphics specific block RMI, ennabling memory related blocks
...
Change-Id: I477adc49b9ee3c8593c193bdc69c0deb4a9726e1
[ROCm/ROCR-Runtime commit: 0e88414f5c
2017-09-25 10:49:22 -05:00
36d1168183
Fix assert in simple_heap.
...
Also add comments to clarify pointer info constraints.
Change-Id: I8d07831a0e953d667c84c96fe53ed07c18ba115c
[ROCm/ROCR-Runtime commit: 476c8e36bf
2017-09-21 00:47:18 -04:00
4c5404b89d
adding hsa_ven_amd_aqlprofile.h to the packaging
...
Change-Id: I3b69396e3cea129106d47be53218213e29de9843
[ROCm/ROCR-Runtime commit: fcaecfee80
2017-09-20 14:40:49 -04:00
595c7130d2
Enable use of CLOCK_MONOTONIC_RAW for post 4.4 kernels.
...
Change-Id: I3c1f27c7e639df5128c36d81f715fa16e6c1cf13
[ROCm/ROCR-Runtime commit: 30fce248c6
2017-09-20 14:28:23 -04:00
c5f1027982
Improve branch elimination in ScopedAcquire.
...
GCC can't reasonably be told that the lock ptr isn't null. Adding a private bool
allows the branch to be eliminated, along with the bool.
Change-Id: I0605d69474d6a6e6951be93c0af1d8caf3f77124
[ROCm/ROCR-Runtime commit: 9dfdce5b3c
2017-09-19 06:08:36 -04:00
3aeea948cb
Add env key to disable 2MB suballocation.
...
Change-Id: Icca3041c3578aa180a656c01aae62f2ad6e8b583
[ROCm/ROCR-Runtime commit: 4275661682
2017-09-19 06:08:36 -04:00
9fd778eec1
Add suballocator for ordinary VRAM allocations smaller than 2MB.
...
Track pointer info for sub 2MB fragment allocations in allocation_map_.
Add fragment support to IPC.
Change-Id: I00cfc2e2fa289aac90a4718c392f9bb056a61a87
[ROCm/ROCR-Runtime commit: 117be0b55a
2017-09-19 06:08:36 -04:00
3dc1b18d54
Use non-RAW version in clock_getres to workaround bug in older kernels
...
Change-Id: Ice0606a42cd7054f0804baf4af3521ffae3b7d50
[ROCm/ROCR-Runtime commit: 87d2df3da3
2017-09-14 13:56:15 -05:00
e150edc48c
Report tools library load failures in debug builds.
...
Change-Id: Ie1ff313e929fc46134e58730a1d370c5d7ace8db
[ROCm/ROCR-Runtime commit: f1a661dedb
2017-08-31 21:32:48 -04:00
b2cb6bca96
Include <errno.h> for EBUSY
...
Change-Id: I9fa3417445866f3ce37af2169f623afa8e92e873
[ROCm/ROCR-Runtime commit: 0f05ef73ac
2017-08-31 07:32:51 -07:00
e18c7098a3
aqlprofile api: adding gfx8 mc counters
...
Change-Id: I84dc06c24b7961dfe665cf7e2ae6cc9ae3b7326b
[ROCm/ROCR-Runtime commit: 09c732c2f9
2017-08-23 15:23:05 -05:00
4ff7a53721
Correct vm_fault signal cleanup.
...
Change-Id: Id2f14b911e3991a76771425bc09f38a613280e6b
[ROCm/ROCR-Runtime commit: 0cb1e8cb35
2017-08-18 22:12:38 -04:00
1fcbd14d89
AqlProfile API, commenting out hanging GFX9 blocks, ATC, GCEA, RPB
...
Change-Id: I411fb33f77c9538ca236b9b6b09c7dfe75220c02
[ROCm/ROCR-Runtime commit: fb4afca8c3
2017-08-15 12:44:00 -05:00
468313f9b5
Simplify pointer info version check.
...
Change-Id: I0ed363f1261ffc041547f313970ca67298ace56c
[ROCm/ROCR-Runtime commit: dec5c52e07
2017-08-12 03:14:39 -04:00
546c179cfc
Initial IPC signal support.
...
Added an API for creating signals with attributes.
Added two APIs for IPC operations on signals.
Initial use of exceptions for error handling.
Add ref counting to signals.
Removed spin loops from signal destructors.
Signals are no longer to be destroyed with delete, use DeleteSignal instead.
Added delete safety to doorbells.
Added secondary hsa_signal_t -> Signal* translation path for IPC enabled signals.
Change-Id: Id59065d002f0c2566b0a9425694da2ed27cb7d7f
[ROCm/ROCR-Runtime commit: c9642cf7af
2017-08-11 18:41:34 -05:00
550547f404
Initial exception support for signals.
...
Also separate signal ABI block allocations from the runtime interface object.
Change-Id: If16763338db664f29163a1348f8f4c38cf0597b2
[ROCm/ROCR-Runtime commit: 2732b18092
2017-08-11 18:41:34 -05:00
ce72ce0c92
adding aqlprofile member to HsaApiTable
...
Change-Id: Id674186dfa2e83295a51f770ccc0400f1cb51a98
[ROCm/ROCR-Runtime commit: 287afd3a52
2017-08-09 16:09:39 -05:00
200067f1be
Adding HSA_API macro to the API method declarations to be consistent with other HSA header files, TCS removing
...
Change-Id: Ic217d3b2bdbb22d3600c5ecaacb7ab53bf26096a
[ROCm/ROCR-Runtime commit: 4824a2db0b
2017-08-08 10:46:12 -04:00
c0c32288ac
aqlprofile block list, explicit numbers assigned, IA removed
...
Change-Id: I9f9358f8e03e13eb81845de2e33dd5f3da27811a
[ROCm/ROCR-Runtime commit: 47322942b3
2017-08-03 11:39:21 -04:00
d224dc274f
aql-profile api: reducing blocks list to compute only and new gfx9 blocks
...
Change-Id: Ib506b82ea407afec4f5d4bcad755d4d57b92e34b
[ROCm/ROCR-Runtime commit: c66f68041c
2017-08-02 12:21:24 -04:00
ea88523fc7
Fix compilation issue reported with GLIBC 2.12 (RHEL 6.9)
...
Change-Id: I770b72ba1d61475a76aa72d0c52ebfb380db6019
[ROCm/ROCR-Runtime commit: 9ee0108e58
2017-07-28 11:11:01 -04:00
4648eb66fd
moving hsa-amd-aqlprofile to ssh://gerritgit/hsa/ec/aqlprofile
...
Change-Id: Ic42752ca41f877db02aa5a5d8d617cd67cce8956
[ROCm/ROCR-Runtime commit: 08d5efe29d
2017-07-14 14:59:42 -05:00
b6d71a8fe6
hsa_ven_amd_aqlprofile.h: include <hsa.h> fix
...
Change-Id: Idfd2fdde112d502d4b4a3365512ec601f7e56a5b
[ROCm/ROCR-Runtime commit: ab67b8511b
2017-07-12 15:43:58 -05:00
dd8804d7ad
Remove use of anonymous member in C builds.
...
Tools/CodeXL will retain older versions of structs if them need them.
Change-Id: I568d7b445778dd575ef71000b4b839300572288e
[ROCm/ROCR-Runtime commit: a0a3587345
2017-07-12 16:40:00 -04:00
41ab59b1e7
Fix queue interception in tools.
...
1. Correct amd::AqlQueue::ExecutePM4 to support interception.
2. Minor fixes to AqlPacket and SoftCP.
3. Minimal change to disable interception of runtime internal queues.
Change-Id: I103fece2ebf9a188d27f01e61221c737405d7253
[ROCm/ROCR-Runtime commit: bc0bd00746
2017-07-12 16:39:43 -04:00
17d0e450cb
Correct handling of slow clocks under linux.
...
Change-Id: I9a1b08d5457caa6739220603bbd37b00febc64d7
[ROCm/ROCR-Runtime commit: 29b5b5c029
2017-07-12 12:49:49 -04:00
19f96afee1
Properly order signal copy agent tagging with copy operation.
...
Change-Id: Ic428c958551279fbea1b2449afba930b82804ede
[ROCm/ROCR-Runtime commit: 3e50adc7ce
2017-07-11 13:10:00 -04:00
cafbebc2a5
Decrement hsa_init ref counter when init fails.
...
Change-Id: If9376344d4b559e601932d070731132c8450104e
[ROCm/ROCR-Runtime commit: c9f0427cb0
2017-07-07 21:21:03 -05:00
ce82829fc1
hsa-runtime integration
...
Change-Id: I48968966ffe164218ebff88d0e3a1268e96bf1dd
[ROCm/ROCR-Runtime commit: 4174f07fd1
2017-07-05 10:55:30 -04:00
7892cc861c
Block list extending
...
Change-Id: Id17efde25fce287296e80f2b37c77b15aa59b561
[ROCm/ROCR-Runtime commit: c533229bc1
2017-06-23 16:37:02 -04:00
9725267667
minor fixes, debug output, comments, using env vars, dead code
...
Change-Id: I08ad73b561709c1818d78a9191c96d6ad141a609
[ROCm/ROCR-Runtime commit: 8618bf7e2c
2017-06-22 18:04:26 -04:00
1351f204d4
Support Perf Cntrs (PMC) and Thread Trace (SQTT) over AQL queues
...
Change-Id: I716b722895d90b46914c31377e791ad602acecc1
[ROCm/ROCR-Runtime commit: 08e0bca567
2017-06-15 12:58:31 -04:00
415027b89f
Revert "Implement memory fault analysis through context save area"
...
This reverts commit 498f3a71885b4df54b10
2017-06-14 14:21:53 -04:00
231d7e8608
GFX8 API
...
Change-Id: I9d0c430e4199f043226c8897f3320a7973cbdeda
[ROCm/ROCR-Runtime commit: 35b376e2ee
2017-06-14 12:24:28 -04:00
498f3a7188
Implement memory fault analysis through context save area
...
When a fatal memory fault occurs the scheduler context-saves all queues
in the process and notifies the runtime through the memory event. The
saved state contains all GPR/LDS data at the moment of the fault.
Retrieve this state and present it to the user if HSA_DEBUG_FAULT is set
to "analyze" and the wavefront caused the fault. If amdgcn-capable objdump
is in the PATH invoke this to disassemble code around the PC.
Queue lifetime is now managed by the runtime to allow querying the
context save state for all active queues.
Change-Id: I6fee662fad1c4f9aa125bf5c53d7d0ea1ab32f95
[ROCm/ROCR-Runtime commit: 75c9506f9d
2017-06-13 23:12:28 -04:00
2cbff7ec5b
Adding HSA extension AMD AQL profile library, see Readme.txt
...
Change-Id: Icbc1e0fb0185642eabbab411a2138ea030d22be8
[ROCm/ROCR-Runtime commit: 25035b8d09
2017-06-13 16:18:06 -04:00
f17287cec4
Adding GFXIP and kernel code object
...
Change-Id: Ieb2dfea8d9e909efac583f541730d77b7d0c9679
[ROCm/ROCR-Runtime commit: da831502ab
2017-06-13 14:58:29 -05:00
79a238eb78
Update hsa_isa_t entries
...
- Add 7.0.2 (consumer hawaii)
- Add 9.0.1 (gfx900 with xnack)
- Add 9.0.2
- Add 9.0.3
Change-Id: I6a07797027c4eaf47038837c5ae51e05b2aba0e4
[ROCm/ROCR-Runtime commit: d98e99949a
2017-06-12 14:34:11 -04:00
f5c06c5911
The fallback path covers not just ARM64, need this for Power as well.
...
Change-Id: I7bbf76f77bd3ac47a0a0987c1e880e23834588e2
[ROCm/ROCR-Runtime commit: a0957bc679
2017-06-07 14:45:29 -05:00
410520a18a
Patch target name in code object for future-proof
...
Change-Id: I6f12b5e5791bd1745ec3ab76d382fad50282e733
[ROCm/ROCR-Runtime commit: cd35fb280a
2017-06-05 19:08:27 -04:00
dfbad5fa07
Enable SDMA on gfx9, disable on gfx8
...
gfx9 has passed qualification. gfx8 stability is under investigation.
Change-Id: Ia72211d47756399ecdfceafeb67c2ab34ebda834
[ROCm/ROCR-Runtime commit: 5db53ceda1
2017-06-02 15:14:14 -05:00
42bb22d0cc
Add preferred agent info to pointer info struct.
...
Lookup blit agent via pointer info in memory_fill.
Change-Id: I02feaf68bb9726858e8cb0ede6bc5f2b3707f5af
[ROCm/ROCR-Runtime commit: c3e2a88ade
2017-05-31 05:16:05 -04:00
1dfe4959e5
remove finalizer usage from image ext
...
Change-Id: I282f02cedce790bf42f07c588fd50e346b9ba665
[ROCm/ROCR-Runtime commit: 77e5b30c41
2017-05-29 20:44:52 -04:00
Sean Keely
Ramesh Errabolu
Chris Freehill
Evgeny
hthangir
Laurent Morichetti
Kenny Ho
Jay Cornwall
Konstantin Zhuravlyov
Qingchuan Shi