Page: Previous  1, 2, 3 ... 5, 6, 7 ... 24, 25, 26  Next

techAdmin
Status: Site Admin
Joined: 26 Sep 2003
Posts: 4128
Location: East Coast, West Coast? I know it's one of them.
Reply Quote
Looks like we need to add xeon support to the new HT detection.

Please paste the contents of: cat /proc/cpuinfo
here and we should be able to get that solved.
Back to top
Ghstryder
Status: Interested
Joined: 08 Sep 2008
Posts: 24
Location: Michigan
Reply Quote
paste.debian.net/26810/

Thanks h2!
Back to top
trash80
Status: Interested
Joined: 03 Nov 2008
Posts: 29
Location: in the middle
Reply Quote
aus9

Mandrake changed their name only on the surface, within the distro the name Mandriva is still used in many places.

This has been thoroughly tested on Mandrake 2007, 2008, and 2009 and has worked flawlessly.

Also, inxi is not made to work ONLY on the newest distro's out there; it is made to work on any distro with the /proc files system inwhich Mandrake 8, 9, and 10 also use. So why remove something made for backward compatibility?

If it works, don't fix it.

Ghstryder,

Thank you for that system, I tried to find someone with a dual socket setup on which to test the new method.

Thanks,

< Edited by trash80 :: Jan 24, 09, 19:52 >

Back to top
techAdmin
Status: Site Admin
Joined: 26 Sep 2003
Posts: 4128
Location: East Coast, West Coast? I know it's one of them.
Reply Quote
Is that truly a dual core, and dual cpu system?

Not hyperthreaded 2 cpus? or anything?

Can you find the actual cpu make and model number so we can compare?

It's giving /proc/cpuinfo the wrong information it appears, unless it's hyperthreaded, then I think it's right.

There were some bugs in inxi as well, but even with those fixed, it doesn't work, because cpuinfo is only showing one true core.

<updated> never mind, it looks like there's an issue with the kernel creating a false /proc/cpuinfo values output, we'll have to take a look at this more closely.
Back to top
trash80
Status: Interested
Joined: 03 Nov 2008
Posts: 29
Location: in the middle
Reply Quote
Ghstryder,

Is that truly a quad core cpu?

/proc/cpuinfo is stating that that is a quad processor system.

Look at the core id's, all 0. A multicore should ID each core per processor. The same id for cores indicates HT.

Also, the processors are numbered 0-3, which means that /proc/cpuinfo is seeing 4 separate processors.

I'd like to see some more examples of other /proc/cpuinfo on quad core Core and Xeon processors.

Thanks,

< Edited by trash80 :: Jan 24, 09, 20:24 >

Back to top
Ghstryder
Status: Interested
Joined: 08 Sep 2008
Posts: 24
Location: Michigan
Reply Quote
This is the processor - there are indeed 2 of them.
This is from the Dell system configuration page.

:: Quote ::
Processor,80546K,3.6G,1M,Xeon Nocona,800,E0

Back to top
trash80
Status: Interested
Joined: 03 Nov 2008
Posts: 29
Location: in the middle
Reply Quote
Ghstryder,

I found this bug report online. Notice the /proc/cpuinfo, the processor is 0 for each one, but each core has a separate id, which is correct behavior. This indicates you have a bug in your kernel.

bugs.launchpad.net/ubuntu/+source/xen-3.2/+bug/262760

Here is another example:
www.linuxquestions.org/questions/linux-general-1/cpu-architecture-explanation-with-proccpuinfo-690607/

This page:
www.phys.uu.nl/~steen/web05a/pentium4.html
states:
:: Quote ::

The Xeon Nocona boast so-called Hyperthreading: with the processor two threads can run concurrently under some circumstances. This may for instance be used for speculative execution of if branches. Experiments have shown that up to 30% performance improvements can be attained for a variety of codes. In practice the performance gain about 3--5%, however.


So there is two HT enabled processors which give the illusion of 4 processors. But the way it is reporting the processors isn't in line with info I have found online. This seems to be some fault in the kernel and xeon processors, after looking at several pages stating bugs in /proc/cpuinfo when xeon is used in linux.

This may be something that cannot be fixed in inxi, but maybe we can find some work around when xeon's are found.

Thank you,
Back to top
trash80
Status: Interested
Joined: 03 Nov 2008
Posts: 29
Location: in the middle
Reply Quote
I see what happens.

ugh

core id's and physical id's don't necessarily numerate in order. Which is a kernel problem. Linus can deal with that. Now I just have to use different logic for core and processor counts.

joy
Back to top
trash80
Status: Interested
Joined: 03 Nov 2008
Posts: 29
Location: in the middle
Reply Quote
:: Ghstryder wrote ::
Once upon a time, I got this output (snipped for the sake of brevity.

:: Code ::
System:    Host gx280-debian running  Linux 2.6.27-8.slh.1-sidux-686 i686 Distro Debian GNU/Linux 5.0
CPU:       Quad Core Intel Xeon cache 1024 KB flags (sse3 nx lm) bmips 28727.4
           Clock Speeds: (1) 3590.929 MHz (2) 3590.929 MHz (3) 3590.929 MHz (4) 3590.929 MHz
 Client Shell inxi 0.8.5


Today, I get this...

:: Code ::
System:    Host gx280-debian Kernel  2.6.28-1.slh.2-sidux-686 i686 (32 bit) Distro Debian GNU/Linux 5.0
CPU:       Single Core HT Intel Xeon cache 1024 KB flags (sse3 nx lm) bmips 7181.88
           Clock Speeds: (1) 3590.941 MHz (2) 3590.941 MHz (3) 3590.941 MHz (4) 3590.941 MHz
Client Shell inxi 0.9.5


For what it's worth. Obviously nothing creates any hardship.
The system has (2) 3.6 GHz dual core Xeons.


Ghstryder

After a nights sleep, and some clear thinking, I can now say that the inxi 0.9.5 output is 100% correct

You don't have dual core processors. The Nocona is a rebadged P4 single core processor with HT.

www.anandtech.com/it/showdoc.aspx?i=2205&p=2

So in the end, it is not a bug in inxi, but expected behavior.

Soon I will implement another piece that will output mult-processor, the logic is there, we just need the output done.

Thanks,
Back to top
Ghstryder
Status: Interested
Joined: 08 Sep 2008
Posts: 24
Location: Michigan
Reply Quote
Thanks trash80!

I had sort of arrived at that, but wasn't sure I was interpreting what I was reading correctly. I know more about HT/Dual Core now. :)

It was those 4 penguins that sent me astray! Thats my story, anyway.
If I had stopped and thought, I would have realized that it was like my P4 HT times 2.
Back to top
Display posts from previous:   
Page: Previous  1, 2, 3 ... 5, 6, 7 ... 24, 25, 26  Next
All times are GMT - 8 Hours