Fixed branch predictor embench generation results

Created by: rosethompson

• begin implemenation of Zicclsm.
• Progress on misaligned load/stores.
• Added file.
• The misaligned load alignment lints.
• Addec ZICCLSM to config files and started on lsu instance.
• At least have the aligner integrated, but not tested.
• Passes lint with some exceptions. Still need to add misaligned store support.
• Progress.
• Finally lints cleanly.
• Aligner is integrated and enabled in rv64gc and passes the regression test; however, there are no new tests.
• rv32gc now also works with the alignment module. Still not tested with misligned access.
• Preemptively fixed the bytemask bug before testing.
• Updated mmu to not generate trap on cacheable misaligned access when supported. Updated tests with David's help.
• First stab at the misaligned test.
• Fixed bugs in misaligned test.
• Progress
• Working through issues with the psill logic.
• Finally the d$ spill works. At least until the next bug. Definitely needs a lot of cleanup.
• Progress. I think the remaining bugs are in the regression test's signature.
• Doesn't yet fully work. Thomas is going to finish debugging while I'm on the RISCV summit next week.
• Enabled Zicclsm in rv64gc.
• Fixed bug which broke the non Zicclsm configs.
• Missed tests.vh.
• Fixed bug in the misaligned access test.
• Fixed spill bugs in the aligner.
• Fixed bug in the Zicclsm test.
• Fixed some more bugs in the Zicclsm signature.
• Fixed all the bugs associated with the signature and the store side of misaligned access. Load misaligned is still causing some issues.
• Found another bug in the RTL's Zicclsm alignment.
• Yay! Zicclsm passes my regression test now.
• Simplification.
• Cleanup.
• Updated buildroot to use kernel 6.6 and added dedicated qemu emulation script.
• Updates to linux config files for sdc.
• Cleanup. Linux makefile wally tracer. probably reduce some complexity here.
• Commented out the arch64priv misaligned load/store tests since we added Zicclsm to the rv64gc config.
• Cleanup and optimization of Zicclsm.
• Fixed bug in uncore updates which broke SDC.
• Reduced Arty A7 clock speed to 20Mhz to support Zicclsm.
• Towards removing the FPGA config file.
• Modified the fpga build script to generate it's own config file rather than use the one in config/fpga.
• Removed fpga config. No longer needed.
• Fixed bugs in the updated fpga synthe script.
• Fixed another bug in the updated script changes.
• Added cbop to to rv32gc.
• Fixed the imperas testbench to be compatible with the config changes.
• Modified the device trees to include all the minor extensions.
• Patched up linux imperas testbench.
• Fixed bug in the btb branch logging. We were only logging branch instructions not all control flow instructions which dramatically skewed the results for sim_bp.
• Fixed second bug in the logger script when branch logging enabled but counter logger not.
• Extended SeparateBranch to support both just branches and all control flow instructions.
• Added btb reference data.
• Updates to btb logger processing.
• Removed the size opt tests from the branch predictor analysis.
• Fixed testbench so it runs with BPRED_LOGGER but not PrintHPMCounters.
• Fixed bugs in paraseHPMC.py