On Thu, Sep 8, 2011 at 9:56 AM, sparsh1 mittal1 <[email protected]>wrote:
> Here is the patch from mailing list of marss posted by Ishwar.
>
> This patch was modified a little to support separate user and kernel stats
collection and current instruction counting is correct.
One more thing: When I see screen display, number of instructions is not
> same as -stopinsns X, but when I see log file, number of instructions, where
> simulation stopped, is same as -stopinsns X. That means code may be correct,
> but my understanding is not complete. Can you explain?
>
> Never rely on the screen output because after a while the screen is not
updated. You have to check the logfile to make sure if the simulation was
correct or not. So, yes the code right now is correct and there is no bug in
instruction counting or in 'stopinsns'.
- Avadh
> ---------------------------
> +++ b/ptlsim/x86/ptlhwdef.h
> @@ -819,6 +819,7 @@ struct Context: public CPUX86State {
> W64 insns_at_last_mode_switch;
> W64 user_*instruction*s_commited;
> W64 kernel_*instruction*s_commited;
> + W64 total_*instruction*s_commited;
> W64 exception;
> W64 reg_trace;
> W64 reg_selfrip;
> --------------------------------------
> +++ b/ptlsim/*core*/ooopipe.cpp
> @@ -2348,6 +2348,7 @@ int ReorderBufferEntry::commit() {
>
> if likely (uop.eom) {
> total_user_insns_committed++;
> + ctx.total_*instruction*s_commited++;
> per_context_ooo*core*_stats_update(threadid,
> commit.insns++);
> thread.total_insns_committed++;
> -----------------------------------
> +++ b/ptlsim/sim/ptl-qemu.cpp
> @@ -723,11 +723,11 @@ void Context::update_mode_count() {
> W64 prev_cycles = cycles_at_last_mode_switch;
> W64 prev_insns = insns_at_last_mode_switch;
> W64 delta_cycles = sim_cycle -
> cycles_at_last_mode_switch;
> - W64 delta_insns = total_user_insns_committed -
> + W64 delta_insns =total_*instruction*s_commited -
> insns_at_last_mode_switch;
>
> cycles_at_last_mode_switch = sim_cycle;
> - insns_at_last_mode_switch = total_user_insns_committed;
> + insns_at_last_mode_switch =total_*instruction*s_commited;
> ------------------------------------
> +++ b/ptlsim/sim/ptlsim.cpp
> @@ -720,6 +720,7 @@ CPUX86State* ptl_create_new_context() {
>
> // Create a new CPU context and add it to contexts
> array
> Context* ctx = new Context();
> + ctx->total_*instruction*s_commited = 0;
> ptl_contexts[ctx_counter] = ctx;
> ctx_counter++;
> -------------------------------------
>
>
> Thanks and Regards
> Sparsh Mittal
>
>
>
>
> On Thu, Sep 8, 2011 at 11:50 AM, avadh patel <[email protected]> wrote:
>
>> Please can you email me that patch so I can confirm that its valid and not
>> fixed in any branch.
>>
>> Thanks,
>> Avadh
>>
>> On Thu, Sep 8, 2011 at 9:48 AM, sparsh1 mittal1
>> <[email protected]>wrote:
>>
>>> Hello
>>> One user submitted a patch for counting instructions in multi-core. It
>>> seems both master and core-models don't have this patch applied.
>>> While doing simulations, I see that total committed instructions are not
>>> same as -stopinsns X for multi-core.
>>>
>>> If the patch is correct, I can apply the patch in my code in my local
>>> computer. Please let me know.
>>>
>>>
>>> Thanks and Regards
>>> Sparsh Mittal
>>>
>>>
>>>
>>> _______________________________________________
>>> 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
