I have seen such problem when you increase your memory latency, for example
to simulate NVMs.
Try increasing "DISPATCH_DEAD_LOCK_COUNTDOWN_CYCLES" in ptlsim/core to a
very high number.

--Ishwan

On Wed, Sep 10, 2014 at 8:47 PM, 高珂 <[email protected]> wrote:

>
> Hi all,
>
> I'm trying to modify the DRAM protocol, and doing some test with marss +
> dramsim2.
>
> some errors occur as follows, my focus is on DRAMSim2, so Marss is not
> that familiar. Who can tell me most possible reasons that happens?
>
> Is read/write, write/read dependency problem, or something else? I'm very
> appreciated with your answers.
>
> [vcpu 3] thread 0: reset thread.last_commit_at_cycle to be before
> redispatch_deadlock_recovery() 993
> [vcpu 3] thread 0: reset thread.last_commit_at_cycle to be before
> redispatch_deadlock_recovery() 993
> [vcpu 3] thread 0: reset thread.last_commit_at_cycle to be before
> redispatch_deadlock_recovery() 993
> Completed         17000 cycles,          4280 commits:     83392 Hz,
> 20995 insns/sec: rip 0000000000401c1c 0000000000401274 ffffffff8100bca0
> 0000000000401293
> [vcpu 3] thread 0: reset thread.last_commit_at_cycle to be before
> redispatch_deadlock_recovery() 993
> [vcpu 3] thread 0: reset thread.last_commit_at_cycle to be before
> redispatch_deadlock_recovery() 993
> [vcpu 3] thread 0: reset thread.last_commit_at_cycle to be before
> redispatch_deadlock_recovery() 993
> [vcpu 3] thread 0: reset thread.last_commit_at_cycle to be before
> redispatch_deadlock_recovery() 993
> Completed         33000 cycles,          4280 commits:     77853 Hz,
>   0 insns/sec: rip 0000000000401c1c 0000000000401274 ffffffff8100bca0
> 0000000000401293
> ....
>
>
>
>
>
>
> _______________________________________________
> http://www.marss86.org
> Marss86-Devel mailing list
> [email protected]
> https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel
>
>
_______________________________________________
http://www.marss86.org
Marss86-Devel mailing list
[email protected]
https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel

Reply via email to