veyepar: Computer Performance Microscopy with SHIM

Author(s):	Xi Yang
Location	D2.193 Percy Baxter
Date	feb Thu 04	Days Raw Files
Start	16:35	First Raw Start	16:16
Duration	0:45:00	Offset	0:18:03
End	17:20	Last Raw End	17:14
Chapters	00:00
Total cuts_time	38 min.

https://linux.conf.au/schedule/30142/view_talk
raw-playlist raw-mp4-playlist encoded-files-playlist host archive tweet mp4 svg png

assets release.pdf Computer_Performance_Microscopy_with_SHIM.json
logs
Admin: episode episode list cut list raw files day marks day marks day image_files

State:

Locked:

clear this to unlock

Locked by:

user/process that locked.

Start:

initially scheduled time from master, adjusted to match reality

Duration:

length in hh:mm:ss

Name:

Video Title (shows in video search results)

Emails:

email(s) of the presenter(s)

Released:

has someone authorised pubication

Normalise:

Channelcopy:

m=mono, 01=copy left to right, 10=right to left, 00=ignore.

Thumbnail:

filename.png

Description:

markdown

Developers and architects spend a lot of time trying to understand and eliminate performance problems. Unfortunately, the root causes of many problems occur at a fine granularity that existing continuous profiling and direct measurement approaches cannot observe. This paper presents the design and implementation of SHIM, a continuous profiler that samples at resolutions as fine as 15 cycles; three to five orders of magnitude finer than current continuous profilers. SHIM’s fine-grain measurements reveal new behaviors, such as variations in instructions per cycle (IPC) within the execution of a single function. A SHIM observer thread executes and samples autonomously on unutilized hardware. To sample, it reads hardware performance counters and memory locations that store software state. SHIM improves its accuracy by automatically detecting and discarding samples affected by measurement skew. We measure SHIM’s observer effects and show how to analyze them. When on a separate core, SHIM can continuously observe one software signal with a 2% overhead at a ~1200 cycle resolution. At an overhead of 61%, SHIM samples one software signal on the same core with SMT at a ~15 cycle resolution. Modest hardware changes could significantly reduce overheads and add greater analytical capability to SHIM. We vary prefetching and DVFS policies in case studies that show the diagnostic power of fine-grain IPC and memory bandwidth results. By repurposing existing hard- ware, we deliver a practical tool for fine-grain performance microscopy for developers and architects.

Comment:

production notes