-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://reviews.gem5.org/r/1927/#review4487
-----------------------------------------------------------



src/mem/SimpleDRAM.py
<http://reviews.gem5.org/r/1927/#comment4205>

    I don't see how the bus width and burst size (in bits/bytes) is of any 
relevance here any longer. Now it is all about the burst length (in beats), is 
it not?



src/mem/SimpleDRAM.py
<http://reviews.gem5.org/r/1927/#comment4206>

    I would change the comment to say, 8 beats across a ...



src/mem/SimpleDRAM.py
<http://reviews.gem5.org/r/1927/#comment4207>

    Once again, the size in bytes is not as relevant here, it is the BL that 
matters.



src/mem/simple_dram.hh
<http://reviews.gem5.org/r/1927/#comment4208>

    ...such that we know when the whole packet is served.



src/mem/simple_dram.hh
<http://reviews.gem5.org/r/1927/#comment4209>

    Perhaps name the variable pktCount here as well?



src/mem/simple_dram.hh
<http://reviews.gem5.org/r/1927/#comment4210>

    Same as above



src/mem/simple_dram.hh
<http://reviews.gem5.org/r/1927/#comment4212>

    Could you add a comment that this is called multiple times on the same 
packet if a system packet is larger than the burst of the memory and that the 
addr is used for the offset within that packet.



src/mem/simple_dram.hh
<http://reviews.gem5.org/r/1927/#comment4211>

    unsigned int



src/mem/simple_dram.hh
<http://reviews.gem5.org/r/1927/#comment4213>

    I still don't understand why this is not const



src/mem/simple_dram.hh
<http://reviews.gem5.org/r/1927/#comment4214>

    And this one :-)



src/mem/simple_dram.hh
<http://reviews.gem5.org/r/1927/#comment4215>

    readBursts?



src/mem/simple_dram.cc
<http://reviews.gem5.org/r/1927/#comment4216>

    I would move this to the initialisation list. Space around '/'



src/mem/simple_dram.cc
<http://reviews.gem5.org/r/1927/#comment4217>

    \n at the end of the DPRINTF string.



src/mem/simple_dram.cc
<http://reviews.gem5.org/r/1927/#comment4218>

    Move to initialisation list for these two.



src/mem/simple_dram.cc
<http://reviews.gem5.org/r/1927/#comment4219>

    \n at the end



src/mem/simple_dram.cc
<http://reviews.gem5.org/r/1927/#comment4220>

    Is the packet ptr still needed here?



src/mem/simple_dram.cc
<http://reviews.gem5.org/r/1927/#comment4221>

    Would this not be easy enough to fix here?



src/mem/simple_dram.cc
<http://reviews.gem5.org/r/1927/#comment4224>

    pktCount > 1, then new BurstHelper here?



src/mem/simple_dram.cc
<http://reviews.gem5.org/r/1927/#comment4222>

    Why involve the absolute address here, and not just divide getSize and 
burstSize?



src/mem/simple_dram.cc
<http://reviews.gem5.org/r/1927/#comment4223>

    wrtie



src/mem/simple_dram.cc
<http://reviews.gem5.org/r/1927/#comment4225>

    Once again I'm not sure of the while loop here. Why not a for look with 
pktCount?



src/mem/simple_dram.cc
<http://reviews.gem5.org/r/1927/#comment4226>

    Alignment here? Worth dealing with that as well I would imagine.



src/mem/simple_dram.cc
<http://reviews.gem5.org/r/1927/#comment4227>

    dram_pkt->burstHelper still points to the burstHelper.


- Andreas Hansson


On June 29, 2013, 10:44 p.m., Amin Farmahini wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://reviews.gem5.org/r/1927/
> -----------------------------------------------------------
> 
> (Updated June 29, 2013, 10:44 p.m.)
> 
> 
> Review request for Default.
> 
> 
> Repository: gem5
> 
> 
> Description
> -------
> 
> This patch gets rid of bytesPerCacheLine parameter and makes the DRAM 
> configuration separate from cache line size. 
> Instead of bytesPerCacheLine, I define a parameter for DRAM called 
> burst_length. The burst_length parameter shows the length of a DRAM device 
> burst in bits.
> Also, I replace lines_per_rowbuffer with device_rowbuffer_size to improve 
> code portablity.
> 
> Updates: 
> - a burst length in beats for each memory type.
> - an interface width for each memory type.
> - the memory controller model is extended to reason about "system" packets vs 
> "dram" packets and assemble the responses properly. It means that system 
> packets larger than a full burst are split into multiple dram packets.
> 
> 
> Diffs
> -----
> 
>   src/mem/SimpleDRAM.py UNKNOWN 
>   src/mem/simple_dram.hh UNKNOWN 
>   src/mem/simple_dram.cc UNKNOWN 
> 
> Diff: http://reviews.gem5.org/r/1927/diff/
> 
> 
> Testing
> -------
> 
> None
> 
> 
> Thanks,
> 
> Amin Farmahini
> 
>

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

Reply via email to