Lerking/nanoBench
1
0
Fork 0
mirror of https://github.com/andreas-abel/nanoBench.git synced 2026-01-28 06:40:09 +01:00
Code Issues Packages Projects Releases Wiki Activity
179 Commits 1 Branch 0 Tags
master
Commit Graph

179 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Andreas Abel
57fab08aad Merge pull request #42 from iwereth/master 
`native_write_msr` compatibilty for kernel 6.16+
 2026-01-26 00:54:39 +01:00
Andreas Abel
9c1f659433 Minor improvements 2026-01-26 00:53:23 +01:00
Andreas Abel
96992fe7c8 Support for RDPRU 
Fixes #44
 2026-01-25 19:15:32 +01:00
Andreas Abel
4802651ad6 Merge pull request #43 from GuillaumeDIDIER/master 
[minor] Fix semi-colon in python
 2026-01-25 17:47:44 +01:00
Andreas Abel
3eb002078a Add a dependency-breaking instruction for false flag dependencies 
Fixes issue #34
 2026-01-24 23:23:43 +01:00
Andreas Abel
110c563f7d Additional TP experiments 2026-01-24 19:39:43 +01:00
Andreas Abel
723c3ae8a2 cpuBench support for ARL-E 2026-01-18 20:00:06 +01:00
Andreas Abel
f752bfeb6c Add experiments with clean upper 32 bits for 32/64-bit GPR operands 2026-01-17 21:17:08 +01:00
Andreas Abel
cea9344dd6 Add an additional latency experiment 
This makes issues like #29 less likely to occur.
 2026-01-09 23:25:17 +01:00
Andreas Abel
ca271bea41 Add support for {rd,wr}{fs,gs}base 
Fixes #30
 2025-11-23 22:32:46 +01:00
Guillaume DIDIER
da0d29434e [minor] Fix semi-colon in python 2025-11-18 08:59:47 +01:00
Andreas Abel
eaf5306080 PTWRITE support 
Fixes #41

Note that PTWRITE potentially has variable throughput, depending on IA32_RTIT_STATUS. This is currently not handled.
 2025-11-16 21:33:09 +01:00
Andreas Abel
884f82742e Initial support for Arrow Lake 2025-11-15 23:40:47 +01:00
Deepak Sharma
78fde5bc68 native_write_msr compatibilty for kernel 6.16+ 
Since kernel 6.16 release, we have the patch https://lore.kernel.org/all/20250427092027.1598740-14-xin@zytor.com/ merged mainline with the commit `785cdec46e92`

It introduces some changes to definition of macros associated with msr operations

Change `native_write_msr` arguments to support this change, along with proper version checks
 2025-11-16 02:31:34 +05:30
Andreas Abel
3980e61377 cpuBench support for EMR, MTL, and ZEN5 2025-06-08 21:39:09 +02:00
Andreas Abel
b3191d26c0 Workaround for https://sourceware.org/bugzilla/show_bug.cgi?id=32813 2025-06-08 21:35:50 +02:00
Andreas Abel
08e351b7fb configs for EMR, MTL, and ARL 2025-05-18 16:22:49 +02:00
Andreas Abel
c7bce4bebc update CPUID 2025-05-17 20:48:06 +02:00
Andreas Abel
217c2c3307 compatibility with recent kernels 2025-03-04 00:43:51 +01:00
Andreas Abel
faf75236ca ranges 2023-03-12 13:48:05 +01:00
Andreas Abel
6fa0df0469 extended asm syntax 2023-01-09 17:25:41 +01:00
Andreas Abel
a7c8989664 Merge pull request #28 from bjzhjing/add-kernel-mode-fix 
fix kernel mode crashing in Ubuntu 22.04
 2022-12-10 20:09:24 +01:00
Cathy Zhang
e13f0aee33 fix kernel mode crashing in Ubuntu 22.04 
While build nanoBench kernel module in Ubuntu 22.04, gcc is with
-mfunction-return=thunk-extern as default option. According to chapter
6.1.1 JMP2RET in the following reference:
https://www.amd.com/system/files/documents/\
technical-guidance-for-mitigating-branch-type-confusion.pdf
all 'ret' instructions are consolidated into a single piece of code.
Instead of functions ending with a 'ret' instruction, they instead end
with "jump __x86_return_thunk".

Since a 'jmp' instruction is provided instead of 'ret' at the end of
each function, it cause functions like create_runtime_code() copy much
more assembler code into runtime_code memory than it should during
runtime. Memory protection fault happens finally while running.

To address the above issue, option -mfunction-return=keep is provided
for kernel mode to overwrite the gcc default behavior in Ubuntu 22.04.
This can ensure function has 'ret' instruction generated.

Signed-off-by: Cathy Zhang <cathy.zhang@intel.com>
 2022-12-06 21:48:56 -08:00
Andreas Abel
c157f74ded improved measurements 2022-12-05 17:11:28 +01:00
Andreas Abel
1ef0dc17dc fixes #26 2022-12-04 00:17:06 +01:00
Andreas Abel
88d021d86e update 2022-11-28 16:32:32 +01:00
Andreas Abel
09e6a3d6fc support for Zen4 2022-11-18 11:32:43 +01:00
Andreas Abel
b0d03d6ad7 memory initialization 2022-08-13 20:56:54 +02:00
Andreas Abel
0dcae1fc1e Zen3 fix 2022-04-28 00:05:02 +02:00
Andreas Abel
d3d4060be3 cycle-by-cycle measurements 2022-01-21 00:12:50 +01:00
Andreas Abel
7f7a7eb53a minor change 2022-01-20 23:21:10 +01:00
Andreas Abel
0ef60f2a59 filter output option 2022-01-20 16:27:58 +01:00
Andreas Abel
15b1ccb275 support for Tremont 2022-01-18 19:01:33 +01:00
Andreas Abel
5bffd03b71 support for Airmont 2022-01-18 02:52:42 +01:00
Andreas Abel
1bbedf2927 minor change 2022-01-17 16:51:56 +01:00
Andreas Abel
86f4a69813 minor fix 2022-01-16 20:40:36 +01:00
Andreas Abel
dfbd89b928 file ownerships 2022-01-15 17:57:28 +01:00
Andreas Abel
a148f95497 helper functions 2022-01-14 20:51:50 +01:00
Andreas Abel
c44e017398 support for Bonnell 2022-01-14 20:22:37 +01:00
Andreas Abel
52e601ebc5 support for Goldmont 2022-01-13 01:14:41 +01:00
Andreas Abel
12381d0324 format specifiers 2022-01-13 01:13:40 +01:00
Andreas Abel
40e764db45 refactoring 2022-01-12 23:21:00 +01:00
Andreas Abel
39f97457c6 improved drain_frontend 2022-01-12 20:33:09 +01:00
Andreas Abel
a1ad0fbf47 consistency 2022-01-12 17:14:49 +01:00
Andreas Abel
3330672330 improved counter configuration 2022-01-12 17:11:46 +01:00
Andreas Abel
73b5ac65e3 FP16 2022-01-11 23:16:44 +01:00
Andreas Abel
6e55106bf5 support for Goldmont Plus 2022-01-08 00:53:37 +01:00
Andreas Abel
9c12de2c70 FP16 2022-01-07 23:40:12 +01:00
Andreas Abel
2a767183a3 minor fixes 2022-01-03 22:26:54 +01:00
Andreas Abel
b156941b23 ADL cache 2022-01-03 22:25:26 +01:00