Re: [gem5-dev] pd-gem5: simulating a parallel/distributed system on multiple physical hosts

Here I’m trying to summarize this long discussion ( I just summarize the points that are discussed in this email thread). 1- Synchronization: using one socket for both communication and synchronization is superior design (multi-gem5). Both pd-gem5 and multi-gem5 use barrier synchronization. 2- Co

Re: [gem5-dev] pd-gem5: simulating a parallel/distributed system on multiple physical hosts

Sure, I am not saying we will get there soon, but I am glad we agree on what is desirable. Actually my reference to SST was not in regard to multi-threading within a host, but as a system that parallelizes the simulation of a single cache-coherent system across multiple hosts. I am not advocating

Re: [gem5-dev] changeset in gem5: dev: add support for multi gem5 runs

I wasn't really ready for this---I thought we were still discussing the pd-gem5 vs. multi-gem5 issues (if slowly). Also, Gabor admitted that async checkpoints are unnecessary, so I think the checkpointing code should be simplified based on that observation. Steve On Wed, Jul 15, 2015 at 5:55 PM

[gem5-dev] changeset in gem5: dev: add support for multi gem5 runs

changeset 5fe05690d03d in /z/repo/gem5 details: http://repo.gem5.org/gem5?cmd=changeset;node=5fe05690d03d description: dev: add support for multi gem5 runs Multi gem5 is an extension to gem5 to enable parallel simulation of a distributed system (e.g. simulation of a pool of

Re: [gem5-dev] Review Request 2845: ruby: Expose MessageBuffers as SimObjects

> On June 11, 2015, 8:42 p.m., Mark Wilkening wrote: > > This patch looks very good to me. Small thing, you might want to consider > > renaming m_buffers_to_free; the name is not really applicable now. Thanks. I changed these names. - Joel ---

Re: [gem5-dev] Review Request 2845: ruby: Expose MessageBuffers as SimObjects

> On June 24, 2015, 4:54 p.m., Joel Hestness wrote: > > configs/ruby/Ruby.py, line 92 > > > > > > Splitting Brad's comments for easier tracking: "Overall, can this code > > be moved into the Switch and Liny .py files rather

Re: [gem5-dev] Review Request 2845: ruby: Expose MessageBuffers as SimObjects

> On June 22, 2015, 5:06 p.m., Brad Beckmann wrote: > > src/mem/ruby/network/MessageBuffer.py, line 39 > > > > > > Why is a default recycle latency not set? > > > > Why is this "Parent.recycle_latency"? Does both

Re: [gem5-dev] Review Request 2968: syscall_emul: standardized file descriptor name and add return checks.

--- This is an automatically generated e-mail. To reply, visit: http://reviews.gem5.org/r/2968/#review6772 --- Ship it! The consistency is nice! Thanks! - Joel Hestness On July 14,

Re: [gem5-dev] Review Request 2969: style: change Process function calls to use camelCase

--- This is an automatically generated e-mail. To reply, visit: http://reviews.gem5.org/r/2969/#review6765 --- Ship it! Thanks for cleaning these up. Below are a couple suggestions, t

Re: [gem5-dev] Deadlock in SystemC Simulation

Hi, Matthias. Things seem to have moved on quite a bit in the mutex behaviour of simulate.cc. The SystemC interworking replaces the simulation loop and hasn't kept up. I think I've recreated your problem and I believe it's related to event queue swapping using ScopedMigration. I believe the pr

[gem5-dev] Cron /z/m5/regression/do-regression quick

* build/ALPHA/tests/opt/quick/se/00.hello/alpha/linux/minor-timing passed. * build/ALPHA/tests/opt/quick/se/00.hello/alpha/linux/o3-timing passed. * build/ALPHA/tests/opt/quick/se/00.hello/alpha/linux/simple-atomic passed. * build/ALPHA/tests/opt/quick/se/00.hello/alpha/linux/simple