derek/gem5
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Commit Graph

  • 5eab6c4b41 Make the notion of a global event tick independent of the actual CPU cycle ticks. This allows the user to have CPUs of different frequencies, and also allows frequencies and latencies that are not evenly divisible by the CPU frequency. For now, the CPU frequency is still set to the global frequency, but soon, we'll hopefully make the global frequency fixed at something like 1THz and set all other frequencies independently. Nathan Binkert 2005-04-11 15:32:06 -04:00
  • 61a20bc32d Merge zizzer.eecs.umich.edu:/bk/m5 into ziff.eecs.umich.edu:/z/binkertn/research/m5/current Nathan Binkert 2005-04-08 21:33:45 -04:00
  • 6fc4c64203 full_system no longer exists as a parameter Nathan Binkert 2005-04-08 21:33:35 -04:00
  • 0d2e816a4c Hand merged a this-> statement for gcc3.4 Ron Dreslinski 2005-04-08 18:26:00 -04:00
  • 2d2f663d3a Add Parameter to only do prefetch calculations on data accesses not instruction accesses Ron Dreslinski 2005-04-08 17:19:56 -04:00
  • dcedd7866e Hand merge Kevin Lim 2005-04-07 16:34:02 -04:00
  • d261ee86c4 Include new CPUs. Kevin Lim 2005-04-07 16:30:40 -04:00
  • 9f2568f966 Support copying memory requests. Nathan Binkert 2005-04-07 00:07:48 -04:00
  • 28ee66cd56 Add support for acking writes with a configurable delay as they are received by the bus bridge. Better Bus debugging. Nathan Binkert 2005-04-06 23:31:31 -04:00
  • 9fead747f5 fix typo in python config stuff Nathan Binkert 2005-04-06 18:00:44 -04:00
  • f3544a13f3 Fix the python NetworkBandwidth conversion function Nathan Binkert 2005-04-06 17:59:31 -04:00
  • 1ee77fb23e formatting Nathan Binkert 2005-04-06 17:58:57 -04:00
  • 060bb32f27 Cleanup diagnostic and error messages for the IDE disk Nathan Binkert 2005-04-06 17:47:32 -04:00
  • 235186859c Better debugging output for the ide controller Nathan Binkert 2005-04-06 17:39:25 -04:00
  • 6d412f63a3 Add TcpPort and UdpPort as python types Nathan Binkert 2005-04-06 17:05:30 -04:00
  • b82bd61946 Move back to vmlinux-latest Nathan Binkert 2005-04-06 17:00:31 -04:00
  • c82562c740 full_system isn't a useful parameter anymore, get rid of it. Nathan Binkert 2005-04-06 16:58:40 -04:00
  • 225730378e Minor fixes for g++ 3.4 Kevin Lim 2005-04-05 16:20:55 -04:00
  • 1b2c81b9d7 Add more prefetcher support. Ron Dreslinski 2005-04-04 16:25:22 -04:00
  • a07340f6cd Merge zizzer:/z/m5/Bitkeeper/m5 into zazzer.eecs.umich.edu:/z/rdreslin/m5bk/prefetcher Ron Dreslinski 2005-04-04 08:34:16 -04:00
  • 4889d8f788 Added support for multiple prefetch address from single access (depth of prefetch) also added the ability to squash some prefetchs to match the GHB technique Ron Dreslinski 2005-04-02 20:36:08 -05:00
  • 46a832eb3b Rework some statistics and add some prefetcher statistics. Also remove an unneeded function call. Ron Dreslinski 2005-04-02 18:40:59 -05:00
  • a72deba708 Fix up Python ParamContext support. Steve Reinhardt 2005-04-02 01:20:47 -05:00
  • 9f2d8346c7 Set the time in the request to proprley calculate latencies Ron Dreslinski 2005-04-01 21:08:03 -05:00
  • 23396b1438 Make sure to calculate the new state properly in coherence protocol Ron Dreslinski 2005-04-01 21:07:08 -05:00
  • fdceb0f00c Some hand merges Ron Dreslinski 2005-04-01 19:39:23 -05:00
  • c21bf8e7ae Another pass at the prefetcher. Now it works with both miss and access reference streams. Reworked how it is instattiated and how it communicates with other cache objects. Ron Dreslinski 2005-04-01 19:26:44 -05:00
  • 9ba57eafc2 Remove a printf that shouldn't be there, Fix some errors with full system config files that wasn't using coherence in MP cases Ron Dreslinski 2005-04-01 18:25:34 -05:00
  • 13608a9b85 Rework the way the prefetcher is used. Now we copy the request from the prefetch queue and into the mq when issued Ron Dreslinski 2005-03-30 15:05:58 -05:00
  • cda8f99a56 First pass at a prefetcher Ron Dreslinski 2005-03-30 04:46:04 -05:00
  • 43a9caa221 expose variables for number of global events per simulated second, millisecond, microsecond, etc. so that the user can explicitly convert between system ticks and time and know what sorts of expensive operations are being used for that conversion. Nathan Binkert 2005-03-29 07:55:44 -05:00
  • 2e2d9b5c35 Merge ktlim@zizzer.eecs.umich.edu:/bk/m5 into zamp.eecs.umich.edu:/z/ktlim2/m5 Kevin Lim 2005-03-28 14:40:02 -05:00
  • a86b95cb18 If we find the mysql libraries on the machine, just compile with mysql and remove the special compile type. Nathan Binkert 2005-03-28 00:41:28 -05:00
  • 1cd15e8e4a style Nathan Binkert 2005-03-25 23:03:31 -05:00
  • 40bab977bc Better handling of latency/frequency parameter types Nathan Binkert 2005-03-25 22:59:29 -05:00
  • 7e1995a29c Better exceptions in python config Nathan Binkert 2005-03-25 22:32:00 -05:00
  • 59f43580bc Update so that statistics can be plotted correctly Ron Dreslinski 2005-03-24 15:52:30 -05:00
  • 50269754f9 Simple python cleanup Nathan Binkert 2005-03-24 14:30:00 -05:00
  • c5b63bc4c1 Improvements to send.py to allow the user to specify the jobfile Nathan Binkert 2005-03-24 12:25:34 -05:00
  • d10412d565 Improve toBool Nathan Binkert 2005-03-24 12:24:54 -05:00
  • eeff53841a Add Frequency and Latency as new parameter types and use them where we can Nathan Binkert 2005-03-24 12:24:17 -05:00
  • 257be74341 Formatting fixes Nathan Binkert 2005-03-23 22:58:47 -05:00
  • 97064c5216 Add some new config files Ali Saidi 2005-03-23 15:57:38 -05:00
  • 5d85dbd458 Change the colors slightly and text (this was for steve's talk way back when) Ali Saidi 2005-03-23 15:57:10 -05:00
  • a8182f74df update profile code to use shared categories file Ali Saidi 2005-03-23 15:55:09 -05:00
  • 903c984770 Merge zizzer:/bk/m5 into zeep.eecs.umich.edu:/z/saidi/work/m5 Ali Saidi 2005-03-23 15:52:03 -05:00
  • 2be988074c Add nhfsstone benchmark and update the latencies of various Monet things Ali Saidi 2005-03-23 15:51:25 -05:00
  • d4e9175c24 Removing unneccasry dprintf that makes traces nearly impossible to read when busses have alot of devices connected to them Ron Dreslinski 2005-03-23 15:23:32 -05:00
  • 153130e558 First step in fixing up parameter handling. Clean up the way ranges work, more fully support metric prefixes for all integer types, and convert memory sized parameters to the MemorySize type. Nathan Binkert 2005-03-23 13:25:48 -05:00
  • df70e5825d Merge zizzer:/bk/m5 into zeep.eecs.umich.edu:/z/saidi/work/m5 Ali Saidi 2005-03-22 16:32:51 -05:00
  • 68129d3384 Yet another validation test Ali Saidi 2005-03-22 16:32:44 -05:00
  • 2ec918362e Merge ktlim@zizzer.eecs.umich.edu:/bk/m5 into zamp.eecs.umich.edu:/z/ktlim2/m5 Kevin Lim 2005-03-22 15:34:08 -05:00
  • 48e0b9ed4d style Nathan Binkert 2005-03-22 14:51:31 -05:00
  • ac547c6489 Fix a bug introduced with the multidict commit. Nathan Binkert 2005-03-22 14:47:18 -05:00
  • 5f2b3a6e5d clean up python exceptions Nathan Binkert 2005-03-22 14:42:05 -05:00
  • 95865117bc Remove lisa's extra printf Nathan Binkert 2005-03-22 14:41:23 -05:00
  • 3583fb6830 Use the multidict in the python config stuff. Makes code a bit cleaner. Nathan Binkert 2005-03-22 00:53:01 -05:00
  • 62ce65c378 We don't support multiple inheritence very well, so we'll just disable it for real. Nathan Binkert 2005-03-22 00:40:21 -05:00
  • eebe4e7fcc Sort the sim objects in the python output Nathan Binkert 2005-03-22 00:08:54 -05:00
  • fc04f8015f Merge zizzer.eecs.umich.edu:/z/m5/Bitkeeper/m5 into zizzer.eecs.umich.edu:/z/stever/bk/m5-head Steve Reinhardt 2005-03-19 08:47:19 -05:00
  • 634b38fc41 Make each etherdev have it's own pci_dev number Ron Dreslinski 2005-03-18 17:17:35 -05:00
  • cfa4221e19 Merge ktlim@zizzer.eecs.umich.edu:/bk/m5 into zamp.eecs.umich.edu:/z/ktlim2/m5 Kevin Lim 2005-03-18 15:32:53 -05:00
  • c8831cfd36 Hand merged some changes Ron Dreslinski 2005-03-18 14:50:48 -05:00
  • a64dd990a0 Use the correct version of linux Ron Dreslinski 2005-03-18 14:43:06 -05:00
  • fbfea4b34e Reworked the configurations to properly use dual/quad checkpoints Ron Dreslinski 2005-03-18 14:41:29 -05:00
  • c929ee2c10 Byproducts of aborted attempt to refine 'parent' proxy semantics. Mostly cleanup of mpy_importer.mpy_parse(). Steve Reinhardt 2005-03-18 13:34:28 -05:00
  • 3efabb6579 re-implement SplitLIFO so that LIFO now applies to individual sets (LRU-style sets) rather than across the whole partition. Lisa Hsu 2005-03-17 16:08:01 -05:00
  • fe764e1276 nake systems parameterizable in the number of NICS. appropriately hook up new nibridges and busses for this scenario. Lisa Hsu 2005-03-17 15:32:10 -05:00
  • 3d51dc4132 Merge zizzer:/bk/m5 into zed.eecs.umich.edu:/z/hsul/work/m5/pact05 Lisa Hsu 2005-03-17 14:34:03 -05:00
  • 5977324255 allow the call to len on Value proxy. Lisa Hsu 2005-03-17 14:31:08 -05:00
  • 8b9e38c303 Merge zizzer.eecs.umich.edu:/z/stever/bk/m5-head into zizzer.eecs.umich.edu:/z/stever/bk/m5-py Steve Reinhardt 2005-03-16 23:10:22 -05:00
  • 0903013669 Allow proxies to refer to proxies in config files. Steve Reinhardt 2005-03-16 23:10:17 -05:00
  • ab5eb7d455 Make panic work in m5.config Nathan Binkert 2005-03-16 20:32:37 -05:00
  • 901aeb24eb Merge zizzer:/z/m5/Bitkeeper/m5 into zizzer.eecs.umich.edu:/.automount/zazzer/z/rdreslin/m5bk/timing_L1 Ron Dreslinski 2005-03-16 18:31:20 -05:00
  • f2dd820972 No need for this ifdef, since the entire process.hh is surounded by an ifndef FULL_SYSTEM Ron Dreslinski 2005-03-16 18:26:32 -05:00
  • 9d6fcfd314 Make the stack size default to 16MB in eio processes Ron Dreslinski 2005-03-16 18:18:59 -05:00
  • 14c461c93c Fixed the super/parent change fpr splash2 benchmarks Ron Dreslinski 2005-03-16 15:55:44 -05:00
  • 486c446110 Remove DUALCPU environment variable and add NUMCPUS, which will build NUMCPUS worth of CPU's. Default is 1 cpu. Ron Dreslinski 2005-03-16 15:45:53 -05:00
  • edaf99eddf Merge zizzer:/z/m5/Bitkeeper/m5 into zizzer.eecs.umich.edu:/.automount/zazzer/z/rdreslin/m5bk/clean Ron Dreslinski 2005-03-16 15:00:56 -05:00
  • b411e0d4ac Move the creation of the L2 out of the memory base (fix for Ali) and make a wrapper that does it Ron Dreslinski 2005-03-16 14:58:27 -05:00
  • d40a1c6573 Some changes to NSGigE's DPRINTF's that helped when I was doing NAT stuff. Lisa Hsu 2005-03-16 13:55:58 -05:00
  • 01caea6df8 Merge zizzer:/z/m5/Bitkeeper/m5 into zizzer.eecs.umich.edu:/.automount/zazzer/z/rdreslin/m5bk/timing_L1 Ron Dreslinski 2005-03-16 13:49:15 -05:00
  • bc6baa4049 Merge zizzer:/bk/m5 into zizzer.eecs.umich.edu:/.automount/zed/z/hsul/work/m5/pact05 Lisa Hsu 2005-03-16 10:49:42 -05:00
  • d805221839 Forgot to commit run.mpy with last changeset (really belongs there). Steve Reinhardt 2005-03-16 10:39:02 -05:00
  • d97590e29d Merge zizzer:/z/m5/Bitkeeper/m5 into zizzer.eecs.umich.edu:/.automount/zazzer/z/rdreslin/m5bk/timing_L1 Ron Dreslinski 2005-03-16 10:31:22 -05:00
  • df012f26fa Fix the bad addr check to check for allowable addresses in the nxm address space Ron Dreslinski 2005-03-16 10:30:33 -05:00
  • c8538d6a7e Enhancements to python config proxy class. Steve Reinhardt 2005-03-16 00:40:48 -05:00
  • 42753edb3c Add a comment to smartdict.py. Steve Reinhardt 2005-03-15 19:41:51 -05:00
  • 6aaa9a7d63 only increment invalid address loads on misspeculated instructions Ali Saidi 2005-03-15 18:07:46 -05:00
  • 6cf53552dd Merge zizzer:/bk/m5 into zeep.eecs.umich.edu:/z/saidi/work/m5 Ali Saidi 2005-03-15 17:31:24 -05:00
  • caf16a99cc during a cache miss in the simple cpu we were finalizing the trace data too early (before the cache miss completed) and therefore writing freeded memory after the cache miss completed. Ali Saidi 2005-03-15 17:31:18 -05:00
  • 363f02a1fd Merge ktlim@zizzer.eecs.umich.edu:/bk/m5 into zamp.eecs.umich.edu:/z/ktlim2/m5 Kevin Lim 2005-03-15 17:26:14 -05:00
  • 0f8067fbf6 Hard code the SimObject::Params struct Nathan Binkert 2005-03-15 17:11:54 -05:00
  • 7d49146ea7 Merge zizzer.eecs.umich.edu:/z/m5/Bitkeeper/m5 into ziff.eecs.umich.edu:/z/binkertn/research/m5/current Nathan Binkert 2005-03-15 15:26:32 -05:00
  • 3833533a64 Merge ktlim@zizzer.eecs.umich.edu:/bk/m5 into zamp.eecs.umich.edu:/z/ktlim2/m5 Kevin Lim 2005-03-15 15:13:24 -05:00
  • dcd8ac5a22 Reduce the client memory requirement for surge so the nat stuff runs on the pool Nathan Binkert 2005-03-15 14:56:40 -05:00
  • 650306d5e5 merge Lisa Hsu 2005-03-15 14:21:39 -05:00
  • aa8ebb9efa Merge zizzer:/bk/m5 into zeep.eecs.umich.edu:/z/saidi/work/m5 Ali Saidi 2005-03-15 13:38:14 -05:00
  • 7e6d5a0461 config file fixes Ali Saidi 2005-03-15 13:37:39 -05:00
  • c2014cb5ad get rid of issequence and just use the isinstance builtin Nathan Binkert 2005-03-15 13:22:47 -05:00