> I looked a little more at the base dir, and there are a bunch more files are > also unused (at least to the extent that I compiled everything successfully > after I removed them). > R src/base/compression/base.hh > R src/base/compression/lzss_compression.cc > R src/base/compression/lzss_compression.hh > R src/base/compression/null_compression.hh > The last remnant's of Erik's cache compression work... I expect that even if > someone revisits that, they may want to use a different interface or > something, so I'm fine with whacking this. Thoughts?
Fine with me. > R src/base/crc.cc > R src/base/crc.hh > Nate? I wonder why those were there. The history didn't help. Did it have to do with checksum offload or something? The tap device perhaps? Anyway, we can always resurrect it. > R src/base/fifo_buffer.cc > R src/base/fifo_buffer.hh > R src/base/hybrid_pred.cc > R src/base/hybrid_pred.hh > R src/base/predictor.hh > R src/base/res_list.hh > R src/base/sat_counter.cc > R src/base/sat_counter.hh > I think these are all classes that Steve Raasch wrote for the old SS-based > out-of-order model that are no longer used. I think we should definitely > get rid of them. > There are two files (sched_list.hh and timebuf.hh) that are used by some of > the CPU models... these seem rather specialized for the "base" dir, and > arguably could go into "sim" or even "cpu". I'd say whack the old SS stuff and move the other two. Nate _______________________________________________ m5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/m5-dev
