aboutsummaryrefslogtreecommitdiff
path: root/benchmarks/common
AgeCommit message (Collapse)AuthorFilesLines
2017-08-04Improve RVC testAndrew Waterman1-1/+2
Make the page-crossing instruction non-idempotent to detect erroneously executing the first 16 bits of the instruction with garbage MSBs.
2017-05-01Set ELF entry point correctlyAndrew Waterman1-0/+1
2017-04-26Add abort() for benefit of benchmark codeAndrew Waterman1-0/+6
2017-03-29Prohibit relaxing the initial gp generationPalmer Dabbelt1-0/+3
I've added an additional linker relaxation that relaxes two instruction pc-relative sequences to one instruction gp relative sequences when possible. This sequence now optimizes the initial gp generation to mv gp, gp which is obviously bogus. The fix is to disable relaxations when setting up gp, preventing the linker from relaxing away this setup code.
2017-03-29Change the global pointer symbol to __global_pointer$Palmer Dabbelt2-2/+2
This changed a while ago in binutils.
2017-03-25Fix FPU initialization codeAndrew Waterman1-6/+8
2017-03-24Like tests, pass the benchmarks if XLEN disagreesAndrew Waterman1-5/+6
2017-03-22Clean up benchmarks buildAndrew Waterman2-33/+10
2017-03-21Clean up bmarks buildAndrew Waterman2-5/+5
2017-03-21Remove smips/host-debugging cruftAndrew Waterman1-50/+0
2017-01-20Use correct macro to guard floating-point initializationAndrew Waterman1-1/+1
2017-01-20Fix build with glibcAndrew Waterman1-0/+2
2016-12-06avoid non-standard predefined macrosAndrew Waterman1-2/+2
2016-08-29On RV32, zero-extend pointers for HTIFAndrew Waterman1-12/+12
2016-07-11Remove deprecated uarch counters; support RVC for benchmarksAndrew Waterman2-27/+13
2016-07-07Don't use FPU in benchmarks that don't need to use the FPUAndrew Waterman3-20/+12
2016-05-25Keep tohost/fromhost at deterministic addressAndrew Waterman3-7/+15
2016-05-02Stop using tohost/fromhost registersAndrew Waterman1-3/+11
2016-04-30ERET -> xRET; new memory mapAndrew Waterman3-10/+96
For now, we no longer build hex files, because the programs don't start at address 0. This decision will likely be revisited.
2016-04-06Don't use stats register; refer to uarch counters by numberAndrew Waterman1-10/+5
2016-03-14Rework benchmarks to run in M-modeAndrew Waterman4-26/+18
This broadens their utility and lets them use the M-mode counters directly.
2016-03-03WIP on priv spec v1.9Andrew Waterman1-3/+3
2015-07-02Fix RV32 handling of syscall argumentsAndrew Waterman2-63/+65
2015-07-01Add _hard_float check in crt.S.Christopher Celio1-0/+3
2015-05-09Update to privileged architecture version 1.7Andrew Waterman3-18/+8
2015-05-01Make dhrystone report correct-ish numbersAndrew Waterman2-1/+16
2015-04-21Don't rely on reset values of MSTATUS_UA/SAAndrew Waterman2-3/+18
2015-04-21Fix benchmark compilation/execution on RV32Andrew Waterman1-1/+1
2015-03-17Merge [shm]call into ecall, [shm]ret into eretAndrew Waterman2-3/+3
2015-03-12Update to new privileged specAndrew Waterman3-38/+40
2015-01-09Add LICENSEAndrew Waterman3-0/+6
2014-12-15Remove dependence on machine/syscall.hAndrew Waterman1-1/+2
2014-12-13Initialize static TLS for the benchmarksAndrew Waterman3-12/+33
2014-12-12Add more entropy to matrix multiplication inputAndrew Waterman1-0/+8
This will exercise the floating-point units more thoroughly.
2014-12-12Use user stack in supervisor modeAndrew Waterman1-14/+1
This fixes a race condition, as there was only one kernel stack.
2014-12-03Use new toolchain and calling conventionAndrew Waterman2-27/+19
2014-11-07Clean up canonical mt benchmarks and reorganize extra versions in /mt. All ↵Henry Cook1-0/+11
versions support support at least 1/2/4 threads.
2014-11-06Fix build with riscv-gcc version 4.9Andrew Waterman1-0/+1
2014-11-06Fix TLS in benchmarksAndrew Waterman1-0/+3
Linker relaxations were screwing up loading the thread pointer.
2014-09-10Enable interrupts in bmarksChristopher Celio1-6/+5
2014-04-14commit high-performance mm (scalar and vector versions)Yunsup Lee3-18/+14
2014-04-07Add radix sort benchmarkAndrew Waterman2-2/+3
2014-04-03setStats in benchmarks now should set and unset the stats register. Also, ↵Stephen Twigg2-5/+14
removed legacy SET_STATS macro.
2014-03-25Make qsort benchmark more meaningfulAndrew Waterman1-2/+2
Before, we were sorting a sorted array :\
2014-02-11Run benchmarks in user modeAndrew Waterman3-49/+186
2014-02-06Clean up benchmarks; support uarch-specific countersAndrew Waterman5-286/+272
2013-11-25Update benchmarks to new privileged ISAAndrew Waterman5-153/+27
2013-10-10Benchmarks now run in user-mode.Christopher Celio2-2/+66
- Jump to main performed by eret. - Nano trap handler added. - FinishTest refactored to perform SYS_exit, placed in util.h. - Only SYS_exit with test_result=1 results in a passed test. - Any other exceptions/syscalls/test_results end program with FAILED test. - PCR status set to S64/U64 if compiled in 64b.
2013-08-23Reflect changes to ISAAndrew Waterman5-796/+147
Conflicts: isa/Makefile
2013-04-29benchmarks initial commitYunsup Lee8-0/+1379