> On Sept. 21, 2015, 2:57 p.m., Nilay Vaish wrote:
> > util/decode_inst_dep_trace.py, lines 73-87
> > <http://reviews.gem5.org/r/3030/diff/1/?file=48965#file48965line73>
> >
> >     I read the code.  I think there is some inconsistency in the decoding 
> > and the encoding procedures.  Typically, the procedures should look like 
> > inverses of each other.
> >     
> >     * When you are writing addr, size and flags, the code in encoding and 
> > decoding differs.
> >      
> >     * PC and weight are not optional in the ascii trace.  You write some 
> > default values.
> 
> Andreas Hansson wrote:
>     I'd just like to point out that the goal here is not to encode/decode 
> protobuf traces to/from ASCII. The decoder is essentially only there to make 
> the trace human readable (and the encoder to be able to create tiny synthetic 
> examples). There is really no value in having two trace formats...

That's exactly right. The ASCII file format will always output a default (0 for 
PC, 1 for weight). I believe this is more human readbale (was to me) than 
having to figure whether or not this optional value is present. For LOAD or 
STORE (the enums added in http://reviews.gem5.org/r/3089/) make it quick to 
spot the addr, size, flags.


- Radhika


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://reviews.gem5.org/r/3030/#review7228
-----------------------------------------------------------


On Aug. 11, 2015, 9:05 p.m., Curtis Dunham wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://reviews.gem5.org/r/3030/
> -----------------------------------------------------------
> 
> (Updated Aug. 11, 2015, 9:05 p.m.)
> 
> 
> Review request for Default.
> 
> 
> Repository: gem5
> 
> 
> Description
> -------
> 
> This patch adds python scripts to parse a protobuf encoded O3CPU
> elastic trace and convert it to a text file output and vice versa.
> 
> 
> Diffs
> -----
> 
>   util/decode_inst_dep_trace.py PRE-CREATION 
>   util/encode_inst_dep_trace.py PRE-CREATION 
> 
> Diff: http://reviews.gem5.org/r/3030/diff/
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> Curtis Dunham
> 
>

_______________________________________________
gem5-dev mailing list
[email protected]
http://m5sim.org/mailman/listinfo/gem5-dev

Reply via email to