- User Since
- Sep 24 2015, 4:45 AM (129 w, 6 d)
Wed, Feb 28
Fri, Feb 23
The idea was that the WAIT in line 11 should ensure that both initial threads arrived.
But actually the runtime is not initialized before line 11. To fix the race, we need to call into the runtime in a way, that makes both threads initial threads before the SIGNAL in line 9.
Thu, Feb 22
As far as I know, the current implementation is very close to the specification in TR6.
Here we have a tiny diff between the spec and the implementation. I agree that we should not apply this patch if we will roll back the change in the spec.
Feb 19 2018
Use the captured pattern as pointed out by Olga
Feb 17 2018
The initial issue in this patch is resolved by D43195. So removed the additionally printed address for AARCH64, but still allow the testcases to match any printed address.
I applied clang-format on commit.
Feb 14 2018
This behavior is right (p.417,l.5):
Feb 9 2018
Feb 6 2018
Jan 30 2018
Jan 25 2018
NEEDS must be REQUIRES
NEEDS replaced by REQUIRES
It was defined in https://reviews.llvm.org/D42432
@omalyshe do you have any advice, what is needed to make the omp_control_tool symbol available on Windows?
Jan 23 2018
Assuming we have pthread on all platforms but Windows
Jan 17 2018
Still surprised that this is necessary for void function calls, but with that change the tests also succeed for intel compiler.
We discussed to relax the wording in the OpenMP spec, so that any implementation that provides ompt_mutex_impl_unknown with value 0 is complying. This will include #define, enum or "static const" values.
Jan 11 2018
Ah, I think I understand your plan now.
Jan 10 2018
Ok, I think, I now understood the change.
Jan 9 2018
@omalyshe AFAIK this issue is Case#: 03138964
I would have expected, that
printf("%" PRIu64 ": &lock: %lu\n", ompt_get_thread_data()->value, (unsigned long) &lock);
would also fix the issue.
As I understand the issue, this is not a fault in the test code, but the value provided by parallel-begin is wrong? In that case, I suggest to pass NULL for this architecture/compiler.
Jan 8 2018
Jan 2 2018
Updated the README
Updated the README
Dec 24 2017
Removed 32 bit ARM from the list of supported architectures until it is tested.
Dec 23 2017
I added the requirement for hardware architecture support to LIBOMP_HAVE_OMPT_SUPPORT.
This way one would also get an error when manually turing LIBOMP_OMPT_SUPPORT=on on an unsupported architecture.
Dec 22 2017
Dec 21 2017
I think, the address needs to be reset on return. Please include commit 45a3a39 to this patch.
Dec 13 2017
This change is more a performace optimization. (reset the bit in the bitmap to 0)
The callback was set to NULL before and all callback invocation checks for the NULL-pointer.
Dec 8 2017
Dec 7 2017
Dec 5 2017
Nov 29 2017
@omalyshe Which version of icc introduced the ident information for workshare constructs?
Nov 24 2017
Nov 22 2017
@kcc I would still be interested in this change :)
Nov 10 2017
We might want to add another RUN-line for 3, where the first tool returns NULL for ompt_start_tool and only the second tool in the list is initialized.
We can add this test in a later patch
Fix potentially uninitialized value. Thanks, Jonas!
Using Olga's code to only trigger a single warning. Tested and works for me.
Remove warning for fini
Addressing Olgas suggestions
Nov 9 2017
Sections seems reasonable, for the single I would suggest to use the conditional.
Nov 8 2017
Instead of saving the address and printing it in a new team afterwards, you could also use CHECK-DAG and print the address, where you store it.
In that case, you should add another, unrelated match in between (e.g., the thread_begin event)
Nov 5 2017
Subsumed by D38185
Nov 2 2017
Updated diff after merging in master with TR4 patch.
Oct 27 2017
Remove the forward declaration in ompt-internal.h.
Remove unnecessary includes of ompt-internal.h.
Replace all _builtin_expect by the UNLIKELY macro.