Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Added submit_keep_args_alive #1395

Merged
merged 13 commits into from
Oct 24, 2023
Merged

Added submit_keep_args_alive #1395

merged 13 commits into from
Oct 24, 2023

Conversation

oleksandr-pavlyk
Copy link
Collaborator

@oleksandr-pavlyk oleksandr-pavlyk commented Sep 7, 2023 •
edited
Loading

Added dpctl.SyclQueue._submit_keep_args_alive(args, events) that increments reference count of args object (typically a sequence of arguments an asynchronous task is operating on), ensuring that args object is not garbage collected until after events signal that tasks working on these objects complete their execution.

  • Have you provided a meaningful PR description?
  • Have you added a test, reproducer or referred to an issue with a reproducer?
  • Have you tested your changes locally for CPU and GPU devices?
  • Have you made sure that new changes do not introduce compiler warnings?
  • Have you checked performance impact of proposed changes?
  • If this PR is a work in progress, are you opening the PR as a draft?

@oleksandr-pavlyk oleksandr-pavlyk force-pushed the async-ref-count-increment branch 2 times, most recently from 3a92246 to 19b628e Compare September 15, 2023 20:40
@github-actions
Copy link

View rendered docs @ https://intelpython.github.io/dpctl/pulls/1395/index.html

@coveralls
Copy link
Collaborator

coveralls commented Sep 15, 2023 •
edited
Loading

Coverage Status

coverage: 85.763% (+0.06%) from 85.701% when pulling 10722d4 on async-ref-count-increment into 1d57614 on master.

@github-actions
Copy link

Array API standard conformance tests for dpctl=0.14.6dev5=py310ha25a700_9 ran successfully.
Passed: 916
Failed: 84
Skipped: 119

@github-actions
Copy link

Array API standard conformance tests for dpctl=0.14.6dev5=py310ha25a700_10 ran successfully.
Passed: 916
Failed: 84
Skipped: 119

@github-actions
Copy link

Array API standard conformance tests for dpctl=0.14.6dev5=py310ha25a700_12 ran successfully.
Passed: 916
Failed: 84
Skipped: 119

@github-actions
Copy link

Array API standard conformance tests for dpctl=0.14.6dev5=py310ha25a700_13 ran successfully.
Passed: 916
Failed: 84
Skipped: 119

@github-actions
Copy link

Array API standard conformance tests for dpctl=0.14.6dev5=py310ha25a700_15 ran successfully.
Passed: 916
Failed: 84
Skipped: 119

@github-actions
Copy link

Array API standard conformance tests for dpctl=0.14.6dev5=py310ha25a700_16 ran successfully.
Passed: 916
Failed: 84
Skipped: 119

@oleksandr-pavlyk oleksandr-pavlyk force-pushed the async-ref-count-increment branch from 6daca36 to 1b393d4 Compare September 19, 2023 14:09
@github-actions
Copy link

Array API standard conformance tests for dpctl=0.15.0rc1=py310ha25a700_14 ran successfully.
Passed: 916
Failed: 84
Skipped: 119

@oleksandr-pavlyk oleksandr-pavlyk force-pushed the async-ref-count-increment branch 2 times, most recently from 4c8ae59 to bb473da Compare September 22, 2023 02:07
@github-actions
Copy link

Array API standard conformance tests for dpctl=0.15.0rc1=py310ha25a700_30 ran successfully.
Passed: 916
Failed: 84
Skipped: 119

@oleksandr-pavlyk oleksandr-pavlyk marked this pull request as ready for review September 22, 2023 22:35
@oleksandr-pavlyk oleksandr-pavlyk mentioned this pull request Sep 23, 2023
Merged
6 tasks
@oleksandr-pavlyk oleksandr-pavlyk force-pushed the async-ref-count-increment branch from bb473da to 7c51f2f Compare September 23, 2023 21:54
@github-actions
Copy link

Array API standard conformance tests for dpctl=0.15.0rc2=py310ha25a700_17 ran successfully.
Passed: 916
Failed: 84
Skipped: 119

@oleksandr-pavlyk oleksandr-pavlyk force-pushed the async-ref-count-increment branch from 11b92bb to 7c51f2f Compare September 25, 2023 22:21
@github-actions
Copy link

Array API standard conformance tests for dpctl=0.15.0rc2=py310ha25a700_24 ran successfully.
Passed: 916
Failed: 84
Skipped: 119

@github-actions
Copy link

Array API standard conformance tests for dpctl=0.15.0rc2=py310ha25a700_25 ran successfully.
Passed: 916
Failed: 84
Skipped: 119

@oleksandr-pavlyk oleksandr-pavlyk force-pushed the async-ref-count-increment branch from 7c51f2f to 7f79887 Compare September 27, 2023 15:19
@github-actions
Copy link

Array API standard conformance tests for dpctl=0.15.0rc3=py310ha25a700_13 ran successfully.
Passed: 916
Failed: 84
Skipped: 119

@oleksandr-pavlyk oleksandr-pavlyk force-pushed the async-ref-count-increment branch from 7f79887 to e303eaa Compare October 5, 2023 20:54
oleksandr-pavlyk added a commit to oleksandr-pavlyk/async-pipelining-example that referenced this pull request Oct 5, 2023
@oleksandr-pavlyk
Python script adopted from IntelPython/numba_dpex#147
73a0794 
This is an adaptation of pipelining technique shared by @mbecker
in https://github.com/IntelPython/numbda_dpex/issues/147

This is built to work with async-ref-count-increment branch
IntelPython/dpctl#1395 which implements
asynchronous memcpy, asynchronous submit and asynchronous
keep_arg_alve task submission.
@github-actions
Copy link

github-actions bot commented Oct 5, 2023

Array API standard conformance tests for dpctl=0.15.1dev0=py310ha25a700_16 ran successfully.
Passed: 935
Failed: 65
Skipped: 119

@github-actions
Copy link

github-actions bot commented Oct 5, 2023

Array API standard conformance tests for dpctl=0.15.1dev0=py310ha25a700_17 ran successfully.
Passed: 935
Failed: 65
Skipped: 119

@oleksandr-pavlyk oleksandr-pavlyk force-pushed the async-ref-count-increment branch from f822827 to ed67ac8 Compare October 6, 2023 01:51
@github-actions
Copy link

github-actions bot commented Oct 6, 2023

Array API standard conformance tests for dpctl=0.15.1dev0=py310ha25a700_18 ran successfully.
Passed: 935
Failed: 65
Skipped: 119

@oleksandr-pavlyk oleksandr-pavlyk force-pushed the async-ref-count-increment branch from ed67ac8 to 802ead7 Compare October 6, 2023 16:22
@github-actions
Copy link

github-actions bot commented Oct 6, 2023

Array API standard conformance tests for dpctl=0.15.1dev0=py310ha25a700_19 ran successfully.
Passed: 935
Failed: 65
Skipped: 119

@github-actions
Copy link

Array API standard conformance tests for dpctl=0.15.1dev0=py310ha25a700_44 ran successfully.
Passed: 935
Failed: 65
Skipped: 119

@oleksandr-pavlyk oleksandr-pavlyk force-pushed the async-ref-count-increment branch from 5b3643c to cebd7f9 Compare October 14, 2023 20:25
@github-actions
Copy link

Array API standard conformance tests for dpctl=0.15.1dev0=py310ha25a700_53 ran successfully.
Passed: 935
Failed: 65
Skipped: 119

@oleksandr-pavlyk
Added SyclQueue._submit_keep_args_alive method
be5fb99 
Usage:
   q = dpctl.SyclQueue()
   ...
   e = q.submit(krn, args, ranges)
   ht_e = q._submit_keep_args_alive(args, [e])
   ....
   ht_e.wait()
@oleksandr-pavlyk
Used _submit_keep_args_alive in a test
5b5363e
@oleksandr-pavlyk
Do not use async_dec_ref in submit method
0a35e9c 
Instead delegated the task of Python object life-time management
to the user via use of _submit_keep_args_alive method
@oleksandr-pavlyk
Add memcpy_async method
f30a6a2
@oleksandr-pavlyk
Introduce dpctl.SyclQueue.submit_async
e7ee1d9 
The SyclQueue.submit has become synchronosing, although
it still returns a SyclEvent (with exectuion_status always complete)
@oleksandr-pavlyk
Use pragma once in _host_task_util.hpp
b6dd4d2
@oleksandr-pavlyk
Fixed possible memory leak on error
306ff96
@oleksandr-pavlyk
Add DPCTLQueue_MemcpyWithEvents
67a325d 
This is the copy operation where one can specify list of events the
copy operation requires before start of its execution.

DPCTLQueue_MemcpyWithEvents(
   __dpctl_keep DPCTLSyclQueueRef QRef,
   void *dst,
   const void *src,
   size_t nbytes,
   const DPCTLSyclEventRef *depEvents,
   size_t nDE
)

Uses this function in tests.
@oleksandr-pavlyk
Adds dpctl.SyclQueue.memcpy_async
6949e69 
Also extends `dpctl.SyclQueue.memcpy` to allow arguments to be objects
that expose buffer protocol, allowing `dpctl.SyclQueue.memcpy` and
`dpctl.SyclQueue.memcpy_async` to be used to copy from/to USM-allocation
or host buffer.
@oleksandr-pavlyk
One of the memcpy_async calls must use events
cf9084d
@oleksandr-pavlyk
Make SyclTimer accumulative
48cb54d 
```

In [9]: timer = dpctl.SyclTimer()

In [10]: with timer(q):
    ...:     y = dpt.linspace(1, 2, num=10**6, sycl_queue=q)
    ...:

In [11]: timer.dt
Out[11]: (0.0022024469999450957, 0.002116712)

In [12]: with timer(q):
    ...:     x = dpt.linspace(0, 1, num=10**6, sycl_queue=q)
    ...:

In [13]: timer.dt
Out[13]: (0.004531950999989931, 0.004239664000000001)
```
@oleksandr-pavlyk
Make cdef function except * to channel Python exceptions
0a73764
@oleksandr-pavlyk
SyclTimer.dt return object with named accessors
10722d4 
The object can unpack into a tuple, like before, but it prints
with annotation of what each number means, and provides names
getters.

  with timer(q):
      code

  dur = timer.dt
  print(dur)     # outputs (host_dt=..., device_dt=...)
  dur.host_dt    # get host-timer delta
  dur.device_dt  # get device-timer delta
  hdt, ddt = dur # unpack into a tuple
@oleksandr-pavlyk
Copy link
Collaborator Author

I tested numba_dpex test suite with dpctl from this PR, and all tests passed.

@oleksandr-pavlyk oleksandr-pavlyk force-pushed the async-ref-count-increment branch from cebd7f9 to 10722d4 Compare October 17, 2023 20:57
@github-actions
Copy link

Array API standard conformance tests for dpctl=0.15.1dev0=py310ha25a700_56 ran successfully.
Passed: 935
Failed: 65
Skipped: 119

@oleksandr-pavlyk oleksandr-pavlyk merged commit 26b9c92 into master Oct 24, 2023
26 checks passed
@oleksandr-pavlyk oleksandr-pavlyk deleted the async-ref-count-increment branch October 24, 2023 00:16
@oleksandr-pavlyk oleksandr-pavlyk mentioned this pull request Jan 26, 2024
Merged
6 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@diptorupd diptorupd Awaiting requested review from diptorupd

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@oleksandr-pavlyk @coveralls