[opensuse-factory] package x86info should be removed from current releases: outdated/unmaintained
Looks unmaintained. Not sure, as current release seems undated on website, but prev release (29) was from 2011. Main detail: shows my 7 yr old server cpu as:
x86info
x86info vVERSION Found 12 identical CPUs Extended Family: 0 Extended Model: 2 Family: 6 Model: 44 Stepping: 2 Type: 0 (Original OEM) CPU Model (x86info's best guess): Unknown model. Processor name string (BIOS programmed): Intel(R) Xeon(R) CPU X5660 @ 2.80GHz Total processor threads: 12 This system has 2 tri-core processors with hyper-threading (2 threads per core) running at an estimated 2.80GHz 2.8 is the max MHz, (1.6 is low), but..a "tri-core processor"...really? did they make such things? Anyway, I think lscpu more than replaces the above info: Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Byte Order: Little Endian CPU(s): 12 On-line CPU(s) list: 0-11 Thread(s) per core: 1 Core(s) per socket: 6 Socket(s): 2 NUMA node(s): 2 Vendor ID: GenuineIntel CPU family: 6 Model: 44 Model name: Intel(R) Xeon(R) CPU X5660 @ 2.80GHz Stepping: 2 CPU MHz: 1597.075 CPU max MHz: 2794.0000 CPU min MHz: 1596.0000 BogoMIPS: 5586.28 Virtualization: VT-x L1d cache: 32K L1i cache: 32K L2 cache: 256K L3 cache: 12288K NUMA node0 CPU(s): 0,2,4,6,8,10 NUMA node1 CPU(s): 1,3,5,7,9,11 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday, 11 October 2018 4:57 L A Walsh wrote:
2.8 is the max MHz, (1.6 is low), but..a "tri-core processor"...really? did they make such things?
Not sure about Intel but AMD did. Originally, these were physically four core chips with one faulty core which was disabled. Later, when the process did not provide enough faulty chips, they started to sell fully functional four core chips with one core disabled. Michal Kubeček -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 11/10/2018 13:27, L A Walsh wrote:
Looks unmaintained. Not sure, as current release seems undated on website, but prev release (29) was from 2011.
Main detail: shows my 7 yr old server cpu as:
x86info
x86info vVERSION Found 12 identical CPUs Extended Family: 0 Extended Model: 2 Family: 6 Model: 44 Stepping: 2 Type: 0 (Original OEM) CPU Model (x86info's best guess): Unknown model. Processor name string (BIOS programmed): Intel(R) Xeon(R) CPU X5660 @ 2.80GHz Total processor threads: 12 This system has 2 tri-core processors with hyper-threading (2 threads per core) running at an estimated 2.80GHz
2.8 is the max MHz, (1.6 is low), but..a "tri-core processor"...really? did they make such things?
If you look below 6 cores per socket with hyperthreading would equate to 3 physical cores per socket so both tools give the same info.
Anyway, I think lscpu more than replaces the above info:
Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Byte Order: Little Endian CPU(s): 12 On-line CPU(s) list: 0-11 Thread(s) per core: 1 Core(s) per socket: 6 Socket(s): 2 NUMA node(s): 2 Vendor ID: GenuineIntel CPU family: 6 Model: 44 Model name: Intel(R) Xeon(R) CPU X5660 @ 2.80GHz Stepping: 2 CPU MHz: 1597.075 CPU max MHz: 2794.0000 CPU min MHz: 1596.0000 BogoMIPS: 5586.28 Virtualization: VT-x L1d cache: 32K L1i cache: 32K L2 cache: 256K L3 cache: 12288K NUMA node0 CPU(s): 0,2,4,6,8,10 NUMA node1 CPU(s): 1,3,5,7,9,11
on my i7 both tools still output meaningful info, i'm betting some people have scripts around that parse parts of its info so while its still getting the core info right i'd be reluctant to drop it unless the current maintainer decides to. -- Simon Lees (Simotek) http://simotek.net Emergency Update Team keybase.io/simotek SUSE Linux Adelaide Australia, UTC+10:30 GPG Fingerprint: 5B87 DB9D 88DC F606 E489 CEC5 0922 C246 02F0 014B
On 10/10/2018 11:44 PM, Simon Lees wrote:
On 11/10/2018 13:27, L A Walsh wrote:
Looks unmaintained. Not sure, as current release seems undated on website, but prev release (29) was from 2011.
Main detail: shows my 7 yr old server cpu as:
x86info x86info vVERSION Found 12 identical CPUs Total processor threads: 12 This system has 2 tri-core processors with hyper-threading (2 threads per core) running at an estimated 2.80GHz
If you look below 6 cores per socket with hyperthreading would equate to 3 physical cores per socket so both tools give the same info.
--- Except that a hyperthread is not a CPU. The above claims 2 threads per core, but if you look out output of lscpu, it says their is 1 thread per core.
CPU(s): 12 Thread(s) per core: 1 Core(s) per socket: 6 Socket(s): 2 NUMA node(s): 2
NUMA node0 CPU(s): 0,2,4,6,8,10 NUMA node1 CPU(s): 1,3,5,7,9,11
on my i7 both tools still output meaningful info,
If you only had x86info, how would you know if its output was correct?
i'm betting some people have scripts around that parse parts of its info so while its still getting the core info right...
The only reason I checked out the source website, was because it got a 7-year old processor way wrong. I wouldn't have bothered to check out the website to look for a way to report the bug if had not been so wrong. Hasn't /proc/cpuinfo been in linux for over a decade or two? Yet this program ignores it the correct info in it. Also the website looks like it lapsed and was purchased and turned into an advertising page. When you click on the link in the x86info page for the authors home page, you get taken to an advertising page. There is no contact information for the original author. Since the author's info (his home page/contact info) is gone, I don't think the website is in his possession any more. Not a shining example of a reliable source at this point. I don't think that makes for a good inclusion unless you can find a bunch of people who need it's output -- but it is poorly formatted for use by other tools (unlike lscpu, which, pretty much, mirrors output in /proc/cpuinfo. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 11/10/2018 18:48, L A Walsh wrote:
On 10/10/2018 11:44 PM, Simon Lees wrote:
On 11/10/2018 13:27, L A Walsh wrote:
Looks unmaintained. Not sure, as current release seems undated on website, but prev release (29) was from 2011.
Main detail: shows my 7 yr old server cpu as:
x86info x86info vVERSION Found 12 identical CPUs Total processor threads: 12 This system has 2 tri-core processors with hyper-threading (2 threads per core) running at an estimated 2.80GHz
If you look below 6 cores per socket with hyperthreading would equate to 3 physical cores per socket so both tools give the same info.
--- Except that a hyperthread is not a CPU. The above claims 2 threads per core, but if you look out output of lscpu, it says their is 1 thread per core.
If you look at the intel website its a 6 core processor with hyperthreading, from an operating system perspective at least in userspace a single core cpu with hyperthreading will just be presented as 2 identical cpu's. Its also possible that this CPU / Motherboard combo was actually implemented as 2 3 core CPU's/sockets built into one combined chip. https://ark.intel.com/products/47921/Intel-Xeon-Processor-X5660-12M-Cache-2-...
CPU(s): 12 Thread(s) per core: 1 Core(s) per socket: 6 Socket(s): 2 NUMA node(s): 2
NUMA node0 CPU(s): 0,2,4,6,8,10 NUMA node1 CPU(s): 1,3,5,7,9,11
on my i7 both tools still output meaningful info, If you only had x86info, how would you know if its output was correct?
Because I know the spec's of the CPU's I buy, after checking the output is right the first time id probably only be using it for scripts.
i'm betting some people have scripts around that parse parts of its info so while its still getting the core info right...
The only reason I checked out the source website, was because it got a 7-year old processor way wrong. I wouldn't have bothered to check out the website to look for a way to report the bug if had not been so wrong. Hasn't /proc/cpuinfo been in linux for over a decade or two? Yet this program ignores it the correct info in it.
Also the website looks like it lapsed and was purchased and turned into an advertising page. When you click on the link in the x86info page for the authors home page, you get taken to an advertising page. There is no contact information for the original author.
Since the author's info (his home page/contact info) is gone, I don't think the website is in his possession any more. Not a shining example of a reliable source at this point.
I don't think that makes for a good inclusion unless you can find a bunch of people who need it's output -- but it is poorly formatted for use by other tools (unlike lscpu, which, pretty much, mirrors output in /proc/cpuinfo.
In the openSUSE project the only standards we have for acceptance are that someone is willing to maintain the package which comes down to them fixing or getting fixed major bugs and keeping the package building. Sure there maybe better tools that people can use but if someone is willing to go to the effort of keeping something working then the project generally won't stop it being included. -- Simon Lees (Simotek) http://simotek.net Emergency Update Team keybase.io/simotek SUSE Linux Adelaide Australia, UTC+10:30 GPG Fingerprint: 5B87 DB9D 88DC F606 E489 CEC5 0922 C246 02F0 014B
On Thursday 2018-10-11 12:19, Simon Lees wrote:
Except that a hyperthread is not a CPU.
It is an execution unit, though. And that's what Linux effectively means by CPU - everytime. Because of history.
If you look at the intel website its a 6 core processor with hyperthreading, from an operating system perspective at least in userspace a single core cpu with hyperthreading will just be presented as 2 identical cpu's. Its also possible that this CPU / Motherboard combo was actually implemented as 2 3 core CPU's/sockets built into one combined chip.
It is also possible the kernel just reports it differently to trick its own scheduler - it has been seen before: Architecture: sparc64 CPU op-mode(s): 32-bit, 64-bit CPU(s): 32 On-line CPU(s) list: 0-23 Off-line CPU(s) list: 24-31 Thread(s) per core: 4 Core(s) per socket: 1 (in reality: 6) Socket(s): 6 (in reality: 1) NUMA node(s): 1 Basically, a processor has a number of execution units. The relationship info however (which unit is associated with which "core" or "socket") is kind of optional, and could be faked in different ways, at different levels. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 11/10/2018 13:20, Jan Engelhardt wrote:
It is an execution unit, though. And that's what Linux effectively means by CPU - everytime. Because of history.
It is almost enough to make you miss BogoMIPS, isn't it? -- Liam Proven - Technical Writer, SUSE Linux s.r.o. Corso II, Křižíkova 148/34, 186-00 Praha 8 - Karlín, Czechia Email: lproven@suse.com - Office telephone: +420 284 241 084 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, 2018-10-10 at 19:57 -0700, L A Walsh wrote:
Looks unmaintained. Not sure, as current release seems undated on website, but prev release (29) was from 2011. Why would you think its unmaintained? openSUSE package is ~1 year old git snapshot [1] and upstream is definitely active [2]. We dont drop software just because it has bugs - I would suggest you report your issues on github bugtracker of project.
Cheers Martin 1. https://build.opensuse.org/package/show/Base:System/x86info 2. https://github.com/kernelslacker/x86info/commits/master P.S openSUSE package is apparently incorectly stated to be version 1.30 when it should be 1.30+gitXYZ
On 10/11/2018 3:10 AM, Martin Pluskal wrote:
On Wed, 2018-10-10 at 19:57 -0700, L A Walsh wrote:
Looks unmaintained. Not sure, as current release seems undated on website, but prev release (29) was from 2011. Why would you think its unmaintained? openSUSE package is ~1 year old git snapshot [1] and upstream is definitely active [2]. We dont drop software just because it has bugs - I would suggest you report your issues on github bugtracker of project.
Cheers
Martin
1. https://build.opensuse.org/package/show/Base:System/x86info 2. https://github.com/kernelslacker/x86info/commits/master
P.S openSUSE package is apparently incorectly stated to be version 1.30 when it should be 1.30+gitXYZ
It also states the URL for the RPM as: URL http://www.codemonkey.org.uk/projects/x86info/ As I've already stated, I went to the website to either get a newer version to test, or report the bug. There was no project at that website. If there were not differences between a hyperthread CPU and a real one, why would the kernel have different schedulers for hyperthread-enabled cpu's vs. non-hyperthread enabled. Just to be clear, this is a system with 2 hexa-core processors with hyperthreading disabled and with Non-Uniform Memory. (a "NUMA" machine). At the verify least the cpus are not identical in that they do not have identical access to the entire range of memory. Hyperthreads are turned off the BIOS. With them enabled, non-hyperthread aware programs would think they had 24 processors -- each with half the memory of the current 12. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 12/10/2018 02.29, L.A. Walsh wrote:
On 10/11/2018 3:10 AM, Martin Pluskal wrote:
On Wed, 2018-10-10 at 19:57 -0700, L A Walsh wrote:
Looks unmaintained. Not sure, as current release seems undated on website, but prev release (29) was from 2011. Why would you think its unmaintained? openSUSE package is ~1 year old git snapshot [1] and upstream is definitely active [2]. We dont drop software just because it has bugs - I would suggest you report your issues on github bugtracker of project.
Cheers
Martin
1. https://build.opensuse.org/package/show/Base:System/x86info 2. https://github.com/kernelslacker/x86info/commits/master
P.S openSUSE package is apparently incorectly stated to be version 1.30 when it should be 1.30+gitXYZ
It also states the URL for the RPM as: URL http://www.codemonkey.org.uk/projects/x86info/
As I've already stated, I went to the website to either get a newer version to test, or report the bug. There was no project at that website.
I would suggest then that you create a bugzilla reporting the incorrect URL and asking what is now the correct upstream for the project, as the stated upstream seems dead. -- Cheers / Saludos, Carlos E. R. (from 42.3 x86_64 "Malachite" at Telcontar)
On 12/10/2018 02.29, L.A. Walsh wrote:
It also states the URL for the RPM as: URL http://www.codemonkey.org.uk/projects/x86info/
As I've already stated, I went to the website to either get a newer version to test, or report the bug. There was no project at that website.
I would suggest then that you create a bugzilla reporting the incorrect URL and asking what is now the correct upstream for the project, as the stated upstream seems dead. ---- That's what prompted me to post here and suggest that the package be removed if it was no longer maintained. Someone here seems to
On 10/12/2018 3:43 AM, Carlos E. R. wrote: think that the source is now on github, however, didn't see anything at the old URL connecting it to the new one or vice versa. So I don't know if they are by the same people or what. FWIW, I did post the tool's output as an incident on github and asked how it came to be that the tool would have such wrong information about an 8yro processor when the information was readily available in proc or via lscpu. So far it has gone unanswered. TBH, I don't really care about the tool. The correct information is readily available out of /proc, and I don't have to wonder whether or not it got it right. It seemed like an ideal candidate for pruning. I've asked people on here, what they thought and got pushback by some people -- I don't care one way or other beyond this point. It seems more appropriate for the SuSE maintainer or those fighting for its continued inclusion to verify the owner's source continuity and make sure there are corrections. Thanks for the ideas! Linda -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
participants (8)
-
Carlos E. R. -
Jan Engelhardt -
L A Walsh -
L.A. Walsh
-
Liam Proven -
Martin Pluskal -
Michal Kubecek -
Simon Lees