Bug 200949
| Summary: | [NEW PORT] benchmarks/stress-ng | ||
|---|---|---|---|
| Product: | Ports & Packages | Reporter: | luca.pizzamiglio |
| Component: | Individual Port(s) | Assignee: | Dmitry Marakasov <amdmi3> |
| Status: | Closed FIXED | ||
| Severity: | Affects Only Me | CC: | luca.pizzamiglio |
| Priority: | --- | ||
| Version: | Latest | ||
| Hardware: | Any | ||
| OS: | Any | ||
| Attachments: | |||
Created attachment 157851 [details]
Poudriere output, FreeBSD 10.1, amd 64
Created attachment 157852 [details]
Poudriere output, FreeBSD 10.1, i386
Created attachment 157853 [details]
Poudriere output, FreeBSD 9.3, amd 64
Created attachment 157854 [details]
Poudriere output, FreeBSD 9.3, i386
Created attachment 157855 [details]
Poudriere output, FreeBSD 8.4, amd64
Created attachment 157856 [details]
Poudriere output, FreeBSD 8.4, i386
Won't it be better to make the while patch conditional to 9.x instead? __FreeBSD_version >= 900000 conditionals may also be removed, by the time I commit this 8.4 would reqch its EoL. Hi Dmitry, I simply used the #ifdef pragma because it was easiest to me. But a part of the patch (#if defined(__STDC_IEC_559_COMPLEX__) and logl redefinition) is relevant to 9.x as well. All specific 8.x modifications will be removed, when 8.4 will reach the EoL. Some patches will be upstreamed (binary number are gcc specific) as well. I prefer to leave the patch so, because in the future it will surely change. At the moment I'm on holiday, so I can reply with big latency Created attachment 158909 [details]
Patch recreated following reviewer's suggestions
Adapting the port following the suggestion of the reviewer.
Now the 8.4 modifications are applied as extra-patches if needed.
I've also updated the version of stress-ng to 0.4.12, as I were here.
Created attachment 161374 [details]
Updating the patch adopting the version 0.4.20 and removing the FreeBSD 8 support
I recreated the patch, removing any support to FreeBSD 8 (discontinued), improving the quality of the patch for FreeBSD 9 (using long double math functions implemented in libstdc++) and upgrading to the version to 0.04.20
I hope it could be committed soon
A commit references this bug: Author: amdmi3 Date: Fri Sep 25 15:14:20 UTC 2015 New revision: 397826 URL: https://svnweb.freebsd.org/changeset/ports/397826 Log: This stress test suite will stress a computer system in various selectable ways It was designed to exercise various physical subsystems of a computer as well as various operating system kernel interfaces. WWW: http://kernel.ubuntu.com/~cking/stress-ng/ PR: 200949 Submitted by: luca.pizzamiglio@gmail.com Changes: head/benchmarks/Makefile head/benchmarks/stress-ng/ head/benchmarks/stress-ng/Makefile head/benchmarks/stress-ng/distinfo head/benchmarks/stress-ng/files/ head/benchmarks/stress-ng/files/extrapatch-Makefile head/benchmarks/stress-ng/files/extrapatch-stress-cpu.c head/benchmarks/stress-ng/files/patch-stress-cpu.c head/benchmarks/stress-ng/pkg-descr |
Created attachment 157850 [details] The patch to add the new port stress-ng is a new set of synthetic benchmarks that exercise various physical subsystems of a computer as well as the various operating system kernel interfaces. It provides: over 90 stress tests, over 50 CPU specific stress tests that exercise floating point, integer, bit manipulation and control flow and over 20 virtual memory stress tests.