- discuss - mailman3.cels.anl.gov

(no subject)
by Cumming Benjamin 06 Apr '18

06 Apr '18

Hello, I have started to play around with Argobots, and I have a first (of many) questions: what is the "rank" of an execution stream (i.e. the value set by ABT_xstream_get_rank)? I also found a small bug in the error handling code, which I made a PR for on the GitHub site: https://github.com/pmodels/argobots/pull/50 Cheers, Ben.

3 5

question about stack reuse on dedicated pools
by Phil Carns 23 Feb '18

23 Feb '18

Hi all, I'm trying to debug problem with a service that spawns a new ULT for each incoming client request. The service works correctly, but memory consumption is considerably higher over time if I create a dedicated pool with it's own execution streams to run the ULTs instead of just running them on the main ES. There are some details here: https://xgitlab.cels.anl.gov/sds/margo/issues/40 That specific test example is creating a total of 40,000 ULTs over the course of execution. It doesn't have that many active concurrently, though. The client program is issuing no more than 4 concurrent requests. A ULT will slightly outlive the request handler, but the number of active ULTS isn't growing indefinitely. I can confirm that argobots is freeing them throughout execution if I turn on Argobots logging. I can see in the Argobots source code that when a ULT is freed it doesn't literally free() the stack memory for that ULT; it is kept on a queue: https://github.com/pmodels/argobots/blob/master/src/include/abti_mem.h#L287 If I print the value of num_stacks in Argobots at the above line, I get the following: - when using the default/main ES: no higher than 512 - when explicitly creating a new pools and ES: 40,000 I haven't tracked down the logic to understand why there is a difference here, but is this expected? It looks like Argobots is keeping a much larger number of stacks around (without reusing them?) if I create a new pool and ES to run my ULTs. thanks, -Phil

2 5

help understanding mutex_unlock logic
by Phil Carns 15 Aug '17

15 Aug '17

Hi all, I'm trying to debug a problem with a custom scheduler (https://xgitlab.cels.anl.gov/sds/abt-snoozer/issues/8) To make a long story short, sometimes we can trigger a scenario where unlocking a mutex does not wake up any ULTs that are blocked on it. I've traced something suspicious in ABTI_mutex_wake_de() path leading up to the deadlock, but I'm not sure if this is an Argobots bug or if I just don't understand the logic. Can someone help sanity check this? In the ABTI_mutex_wake_de() invocation just before the hang, the value of num_elem is 1 at this line: https://github.com/pmodels/argobots/blob/master/src/mutex.c#L769 ... but then after checking the high and low priority lists it falls through to here, having not found anything to wake up: https://github.com/pmodels/argobots/blob/master/src/mutex.c#L802 Is that code path supposed to be possible? It seems non intuitive that the count could be > 0 but it can't find a thread to wake up, but there might be a more subtle meaning to num_elem. The particular scheduler that I am debugging will be blocking/sleeping until it finds work to do, so it's important that this path ultimately triggers a pool push or else it won't make progress. That may not be an issue with other schedulers. thanks! -Phil

2 3

how to marshal/serialize an ABT_pool id for network use
by Phil Carns 27 Feb '17

27 Feb '17

Hi all, I'm considering how to build a network service in which incoming requests could potentially be serviced by more than one Argobots pool (ABT_pool). At the most fundamental level, this will look like a single centralized progress loop for the network that decodes the header for incoming messages, and based on that header calls ABT_thread_create() to launch the handler on a particular pool. I'd like the client to be able to control which pool it is targeting. I think this means that I should probably create my own data structure that maps "pool_id -> ABT_pool" so that I have a generic identifier that I can share over the network with other hosts. "pool_id" in this case would just be an int64_t or something similar of a fixed size at any rate. When a request arrives I would decode the header, look up the pool_id to ABT_pool mapping, then call ABT_thread_create() on that pool. Before I implement that, though, I just wanted to check- there isn't any existing mechanism built into Argobots for generating stable IDs for a given ABT_pool, is there? thanks! -Phil

3 5

discuss test2
by Seo, Sangmin 09 Dec '16

09 Dec '16

test2

1 0

another test
by Seo, Sangmin 07 Dec '16

07 Dec '16

test

1 0

Argobots question
by Sangmin Seo 18 Nov '16

18 Nov '16

test again...

2 1

discussion
by Seo, Sangmin 17 Nov '16

17 Nov '16

test

1 0

[bolt-discuss] question
by Seo, Sangmin 17 Nov '16

17 Nov '16

This is a test...

1 1