June 2005 - openSUSE AMD64 - openSUSE Mailing Lists

RE: [suse-amd64] AMD Dual Core & kernel 2.6
by James D. Parra 13 Jun '05

13 Jun '05

> Will the kernel in Suse 9.1 support Dual AMD, Dual Core, processors? >It should work, but no optimizations and the CPUs are not reported >as siblings (this means if you have a license manager that licenses >per CPU it might make you pay too much) So the kernel will only see 2 CPU's not 4? Thanks, ~James

2 1

RE: [suse-amd64] AMD Dual Core & kernel 2.6
by James D. Parra 13 Jun '05

13 Jun '05

On 13 Jun, James D. Parra wrote: > Will the kernel in Suse 9.1 support Dual AMD, Dual Core, processors? It should.. the OS just sees seperate cpus like they were on their own dies.. not like Hyperthreading on the P4 line. Sles9 sp1 box: processor : 3 vendor_id : AuthenticAMD cpu family : 15 model : 33 model name : Dual Core AMD Opteron(tm) Processor 275 stepping : 2 cpu MHz : 2205.027 cache size : 1024 KB physical id : 1 siblings : 2 fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext lm 3dnowext 3dnow pni bogomips : 4407.29 TLB size : 1088 4K pages clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts fid vid ttp egrep 'proc|model name|physical|sibling' /proc/cpuinfo processor : 0 model name : Dual Core AMD Opteron(tm) Processor 275 physical id : 0 siblings : 2 address sizes : 40 bits physical, 48 bits virtual processor : 1 model name : Dual Core AMD Opteron(tm) Processor 275 physical id : 0 siblings : 2 address sizes : 40 bits physical, 48 bits virtual processor : 2 model name : Dual Core AMD Opteron(tm) Processor 275 physical id : 1 siblings : 2 address sizes : 40 bits physical, 48 bits virtual processor : 3 model name : Dual Core AMD Opteron(tm) Processor 275 physical id : 1 siblings : 2 address sizes : 40 bits physical, 48 bits virtual Thanks Mike, this is good to know. ~James

1 0

Re: [suse-amd64] AMD Dual Core & kernel 2.6
by Michael Marion 13 Jun '05

13 Jun '05

On 13 Jun, James D. Parra wrote: > Will the kernel in Suse 9.1 support Dual AMD, Dual Core, processors? It should.. the OS just sees seperate cpus like they were on their own dies.. not like Hyperthreading on the P4 line. Sles9 sp1 box: processor : 3 vendor_id : AuthenticAMD cpu family : 15 model : 33 model name : Dual Core AMD Opteron(tm) Processor 275 stepping : 2 cpu MHz : 2205.027 cache size : 1024 KB physical id : 1 siblings : 2 fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext lm 3dnowext 3dnow pni bogomips : 4407.29 TLB size : 1088 4K pages clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts fid vid ttp egrep 'proc|model name|physical|sibling' /proc/cpuinfo processor : 0 model name : Dual Core AMD Opteron(tm) Processor 275 physical id : 0 siblings : 2 address sizes : 40 bits physical, 48 bits virtual processor : 1 model name : Dual Core AMD Opteron(tm) Processor 275 physical id : 0 siblings : 2 address sizes : 40 bits physical, 48 bits virtual processor : 2 model name : Dual Core AMD Opteron(tm) Processor 275 physical id : 1 siblings : 2 address sizes : 40 bits physical, 48 bits virtual processor : 3 model name : Dual Core AMD Opteron(tm) Processor 275 physical id : 1 siblings : 2 address sizes : 40 bits physical, 48 bits virtual -- Mike Marion-Unix SysAdmin/Staff Engineer-http://www.qualcomm.com Like the man said "Nothing is foolproof, because fools are so ingenious"

1 0

AMD Dual Core & kernel 2.6
by James D. Parra 13 Jun '05

13 Jun '05

Hello, Will the kernel in Suse 9.1 support Dual AMD, Dual Core, processors? Thanks, James

2 1

Re: [suse-amd64] Tyan S2881 network problem
by Michael Marion 13 Jun '05

13 Jun '05

On 13 Jun, Brad Dameron wrote: > It is not the BC57xx chipset in general. It is the implementation of it > on the Tyan S2881 early versions of the motherboard. They have since > corrected it and you can get a replacement motherboard from them. There's a similar problem we've been seeing with HP DL360's (I think) and (definitely) DL385s and the 5704 chipset where the NIC just shuts down. Downing then re-upping the interface brings it back online. Supposedly a new firmware fixes it, though we had at least one updated box hang the same way again on Fri. -- Mike Marion-Unix SysAdmin/Staff Engineer-http://www.qualcomm.com "Only wimps use tape backup: _real_ men just upload their important stuff on ftp, and let the rest of the world mirror it ;)" (Linus Torvalds, about his failing hard drive on linux.cs.helsinki.fi)

1 0

Tyan S2881 network problem
by Ed Sherman 13 Jun '05

13 Jun '05

Hello all, I'm trying to install SLES9 on a Tyan K8SR (S2881) The install goes okay but when the system start up I get: eth0 (DHCP) . . . . . no IP address yet... backgrounding. .......... waiting Same thing with eth1 (The board has two NICs) I can't connect to another machine via network. I can only ping localhost. I'm wondering if others have had this problem on this same board or if the network config goes well on the other Tyan K8SR (S2881) out there. I've searched for this on the net and found other people having similar problems but not for this particular motherboard or network interface (Broadcom 5704). If you've installed with this board and it went well please let me know. If you had problems with networking on this board and solved it (or didn't) please let me know if you have a solution. I'm starting to suspect a bad network interface on the board. Thanks in advance, Ed

3 3

RE: [suse-amd64] Tyan S2881 network problem
by Black, Alain 13 Jun '05

13 Jun '05

It sounds like your options are to replace the mother boards (more than likely cost prohibitive to say the least) or to get PCI cards for your network needs. -Alain. -----Original Message----- From: Joe Georger [mailto:jgeorger@ll.mit.edu] Sent: Monday, June 13, 2005 12:44 PM To: Black, Alain Cc: Brad Dameron; suse-amd64(a)suse.com Subject: Re: [suse-amd64] Tyan S2881 network problem Our files are typically 2 GB each. I think Tyan is using the non-C variant on their newer, PCI-E/nforce Pro boards. In googling I have seen a post of two of people saying the BCM5704C dies under anything more than a moderate load. Joe Black, Alain wrote: >Joe, > >What do you consider large files? > >I've got a 2 servers with MSI mother boards with the BCM5704 (don't know >what the C difference is) chip and we nfs/sftp/rcp multi gig files. >Nothing larger than 12 GB at this point though. We are quite pleased >with their performance. > >This is both under SLES8SP3 and SLES9. > >-Alain > >-----Original Message----- >From: Joe Georger [mailto:jgeorger@ll.mit.edu] >Sent: Monday, June 13, 2005 12:17 PM >To: Brad Dameron >Cc: suse-amd64(a)suse.com >Subject: Re: [suse-amd64] Tyan S2881 network problem > >We've got 7 of these units. On one the ethernet is dead. Some of them >have qlogic cards and are serving 2-3 3TB raids. When we try to move >copy large files over nfs or rcp, the machines like to lock up. We have >gone from Suse 9.1 Pro to 9.2 Pro to SLES9 and nothing is staying up. > >Is the BCM5704C really that bad? I can't blame the OS because I just >can't believe SLES9 would lock up like that. > >Joe > >

1 0

RE: [suse-amd64] Touchpad Issues in 9.3
by Black, Alain 13 Jun '05

13 Jun '05

-----Original Message----- From: Andi Kleen [mailto:ak@suse.de] Sent: Saturday, June 11, 2005 6:28 AM To: Black, Alain Cc: Barry Premeaux; suse-amd64(a)suse.com Subject: Re: [suse-amd64] Touchpad Issues in 9.3 On Fri, Jun 10, 2005 at 09:03:35AM -0700, Black, Alain wrote: > I have a similar issue with my desktop machine running SuSE 9.3 32bit > that's hooked up to a KVM. When I switch back to the SuSE desktop the > mouse goes haywire (large movements across the screen and random mouse > clicks when I do not click the mouse buttons). The PC is a Dell > Optiplex GX270. > > Then I get "lost synchronization, throwing 3 bytes away", sometimes it's > 2 bytes and other times 1 byte. I never see the driver resynched > message though. > > My workaround (part of which will not work for a touch pad) is to move > the mouse about a quarter of an inch, which throws the cursor all over > the screen, and then let it sit for a couple of seconds. When that > doesn't work, I pick the mouse up until it can't read a surface below it > and put it back down. > Boot with psmouse.proto=bare. That fixes all problems, but unfortunately also disables the scrollwheel :/ -Andi -- Andi, Thanks, that option works for me. Just a note: I do not have this problem when attached to KVM's from Avocent, only with the less robust SoHo model from Belkin that I'm using at my desk. -Alain

2 1

RE: [suse-amd64] Mobos/Chipsets for Athlon64 3200+ &SuSE Linux 9.x
by Black, Alain 13 Jun '05

13 Jun '05

I currently use an ATI x800 XT PE on that mother board. ATI has been getting better with their linux drivers. Their current release of drivers 8.14 even has an actual gui installer. Nvidia currently only supports up to the 6200 cards. This update to include the 6200's came out earlier this month from Nvidia. SuSE does a very nice job of making both vendors drivers simpler to install by repackaging them. Both vendors (ATI and Nvidia) recommend using the SuSE install process in the read me for their drivers. -Alain. -----Original Message----- From: Eugene Chu [mailto:chu@tes-mail.jpl.nasa.gov] Sent: Friday, June 10, 2005 12:49 PM To: Black, Alain; kus(a)free.net; suse-amd64(a)suse.com Subject: RE: [suse-amd64] Mobos/Chipsets for Athlon64 3200+ &SuSE Linux 9.x Alain Black wrote: >I have an MSI K8N Neo2 Platinum that I am quite pleased with. I use >the Parallel ATA as well at SATA. The only piece that I have not tried >yet is the firewire as I haven't had any digital editing to do since I >hooked this system up. I'm glad to hear that, as I am also looking at that board for my next computer. Mikhail Kuzminsky wrote: >I don't want to use Nvidia chipsets because of potential problems >w/graphics. OK, there is VIA K8T800 Pro chipset. I think you might be confusing the nvidia mobo chipsets with their graphics chipsets. The K8N Neo2 Platinum that Alain referred to above uses a nvidia gforce3 ultra I/O chip, and works quite well on various flavors of Linux, based on all the reports I've seen of it. Plus it has the built-in Gig-E interface, which can give you full Gig-E performance without sapping up bandwidth on the PCI bus. I do not believe the VIA K8T800 Pro has this feature. You don't need to use a nvida graphics card with this mobo; something from ATI should work just as well. Nvidia used to claim to have good support for their graphics chips on Linux systems. I'm not sure what the situation is now, so I would like to learn that as well. eyc

2 1

Re: [suse-amd64] number of CPUs for SMP kernels
by Mikhail Kuzminsky 13 Jun '05

13 Jun '05

In message from "Andi Kleen" <ak(a)suse.de> (Sat, 11 Jun 2005 03:15:58 +0200): >On Fri, Jun 10, 2005 at 05:19:13PM -0700, Black, Alain wrote: >> Does that count physical processors or cores on the processors? > >local APIC IDs, cores or even HT threads. > >Remember he asked for ancient distributions. Yes, you are right about "ancient" :-) The reason of question was just the possibility to use this versions of SuSE w/2 chips*2 cores = 4 cores and modern "dual" (4 cores) server motherboards. Do I understasnd you correctly, that this SuSE versions will (at least in theory) work w/4 cores ? Yours Mikhail >Modern ones (including >SLES9-SP2) >have these limits extended, however there is no supported AMD setup >right now with that many CPUs. > >> Currently AMD has dual core CPU's on the market, they are supposed >>to >> release quad core CPU's sometime in Q1 2006. >> >> Would we then be restricted to 2 quad core CPU's? > >8 APIC ID is four socket dual core or 8 socket single core. > >-Andi > >-- >Check the List-Unsubscribe header to unsubscribe >For additional commands, email: suse-amd64-help(a)suse.com > >

2 1