Re: AMD K6-2 / Bogomips problem
Hi! On 19-Jan-99, Peter Bartosch took time to write : bogomips should only depend on hardware no ? only on cpu (and clock) that's what i thought too, but it doesn't seem true in my case as with exactly the same hardware and no change in bios i have 700 bogomips with kernel 2.2.0pre7 and 350bogomips with kernel 2.0.36 compiled either as 586 or as 686 they are all compiled with same options (as much as possible) i've read somewhere, that this lies on the size of the cpu's cache (1st level) this is IIRC in K6-chips twice as big as in pentiums, but if i think over it this couldn't be true/realistic until next mail B-) Peter -- :~~ [EMAIL PROTECTED] ~: : student of technical computer science : : university of applied sciences krefeld (germany) : ~~ FD314F21 C7 AE 2F 28 C1 33 71 77 0D 77 CD 6E 58 E9 06 6B
Re: AMD K6-2 / Bogomips problem
On Thu, 21 Jan 1999, Peter Bartosch wrote: On 19-Jan-99, Peter Bartosch took time to write : bogomips should only depend on hardware no ? only on cpu (and clock) that's what i thought too, but it doesn't seem true in my case as with exactly the same hardware and no change in bios i have 700 bogomips with kernel 2.2.0pre7 and 350bogomips with kernel 2.0.36 compiled either as 586 or as 686 they are all compiled with same options (as much as possible) i've read somewhere, that this lies on the size of the cpu's cache (1st level) this is IIRC in K6-chips twice as big as in pentiums, but if i think over it this couldn't be true/realistic Actually, the main critique of the real MIPS calculation is that it is cache-dependent because it is a small loop. While I know nothing of the bogomips calculation, I assume it's similar. The problem is that the entire piece of code that calculates this fits inside the cache. Pentiums (and MMX) came standard w/ 512k, so I doubt the loop would be bigger than that, so I don't think doubling the cache size would help any. (Just celerons would really get bad numbers) :) -Dan
Re: AMD K6-2 / Bogomips problem
On 19-Jan-99, Peter Bartosch took time to write : bogomips should only depend on hardware no ? only on cpu (and clock) that's what i thought too, but it doesn't seem true in my case as with exactly the same hardware and no change in bios i have 700 bogomips with kernel 2.2.0pre7 and 350bogomips with kernel 2.0.36 compiled either as 586 or as 686 they are all compiled with same options (as much as possible) Patrick
Re: AMD K6-2 / Bogomips problem
Hi! i have : processor : 0 cpu : 586 model : AMD-K6(tm) 3D processor vendor_id : AuthenticAMD stepping: M difference !!! what is it ? it's a production-revision! higher stepping - newer cpu because of newer production-masks (and hopefully with less errors) i have two 64Mo chips 100MHz coming from different manufacturers (could that be the cause ?) will bogomips change if memory changes ? no it won't, bogomips depend only on your cpu (-registers) I have found in the past on a couple of occasions I have had ludicrously low bogomips figures and have re-compiled the kernel (not even changing the settings!) and the rate has gone up the next time i have re-booted. Only ever happened with a RedHat system - don't know why. isn't that strange ? don't worry about it, bogomips are bogus, they don't say anything since the bogomips are equal to your clock frequency (or to the double clock freq with some (all) K6) maybe them could be used as comparision-argument, but this was in pre P5 times bogomips should only depend on hardware no ? only on cpu (and clock) just my 0.02 euro until next mail ;) Peter -- :~~~ [EMAIL PROTECTED] ~~: : student of technical computer science: : university of applied sciences krefeld (germany) :
Re: AMD K6-2 / Bogomips problem
[EMAIL PROTECTED] wrote: Hi, I think i have a problem with one of my new K6-2 box. I have two boxes : same motherboard (ASUS P5A) , same amount of memory (128Mo). Same output of /proc/cpuinfo for both *except* for bogomips. One is a 350MHz and it gives me ~350 Bogomips the other is a 400MHz and it gives me ~800 Bogomips I think there is a problem. What do you think ? What should i check first ? Interesting. I too am using a K6-2. My /proc/cpuinfo reads: processor : 0 cpu : 586 model : 8 vendor_id : AuthenticAMD stepping: unknown fdiv_bug: no hlt_bug : no f00f_bug: no fpu : yes fpu_exception : yes cpuid : yes wp : yes flags : fpu vme de pse tsc msr mce cx8 mmx bogomips: 699.60 My motherboard is an AOpen AX59Pro. I'm not over-clocked however. I'm also using the kernel I installed with too. -- __ _ Mark Wagnon -o) / / (_)__ __ __ Chula Vista, CA /\\/ /__/ / _ \/ // /\ \/ / mailto:[EMAIL PROTECTED] _\_v/_/_//_/\_,_/ /_/\_\
Re: AMD K6-2 / Bogomips problem
On Sun, Jan 17, 1999 at 11:13:41AM +0100, [EMAIL PROTECTED] wrote: I think i have a problem with one of my new K6-2 box. I have two boxes : same motherboard (ASUS P5A) , same amount of memory (128Mo). Same output of /proc/cpuinfo for both *except* for bogomips. One is a 350MHz and it gives me ~350 Bogomips the other is a 400MHz and it gives me ~800 Bogomips I think there is a problem. What do you think ? I just bought a P5A and 350MHz K6-2 last week. I get almost exactly 700 bogomips (699 or 701 or so). Make sure the bus speed is 100MHz and that cache is all turned on in the BIOS. What does /proc/cpuinfo say about the CPU clock? (Only in Linux 2.1 2.2). Nice system, wish I had faster disks to match it. Hamish -- Hamish Moffatt VK3TYD [EMAIL PROTECTED], [EMAIL PROTECTED] Latest Debian packages at ftp://ftp.rising.com.au/pub/hamish. PGP#EFA6B9D5 CCs of replies from mailing lists are welcome. http://hamish.home.ml.org
Re: AMD K6-2 / Bogomips problem
On Sun, Jan 17, 1999 at 11:13:41AM +0100, [EMAIL PROTECTED] wrote: One is a 350MHz and it gives me ~350 Bogomips the other is a 400MHz and it gives me ~800 Bogomips I think there is a problem. What do you think ? No. What should i check first ? Read the Bogomips mini-Howto. /usr/doc/HOWTO/mini/BogoMips.gz Marcus -- Rhubarb is no Egyptian god.Debian GNU/Linuxfinger brinkmd@ Marcus Brinkmann http://www.debian.orgmaster.debian.org [EMAIL PROTECTED]for public PGP Key http://homepage.ruhr-uni-bochum.de/Marcus.Brinkmann/ PGP Key ID 36E7CD09
Re: AMD K6-2 / Bogomips problem
On 17-Jan-99, Marcus Brinkmann took time to write : One is a 350MHz and it gives me ~350 Bogomips the other is a 400MHz and it gives me ~800 Bogomips I think there is a problem. What do you think ? No. i think there is, see below What should i check first ? Read the Bogomips mini-Howto. /usr/doc/HOWTO/mini/BogoMips.gz besides the fact it seems outdated (1997-12-13), it says like me: AMD K5/K6 clock * (2.00 plusminus 0.010) 11.1 and it lists K6 at 166Mhz having already ~330Bogomips ! so my AMD K6-2 at 350Mhz should have 700Bogomips isn't it ? so next part of my question: what should i check ? bios settings ? (the howto is not precise enough imho) Patrick
Re: AMD K6-2 / Bogomips problem
On Sun, Jan 17, 1999 at 03:34:42PM +0100, [EMAIL PROTECTED] wrote: besides the fact it seems outdated (1997-12-13), it says like me: AMD K5/K6 clock * (2.00 plusminus 0.010) 11.1 and it lists K6 at 166Mhz having already ~330Bogomips ! so my AMD K6-2 at 350Mhz should have 700Bogomips isn't it ? I don't know what -2 means at K6-2. so next part of my question: what should i check ? bios settings ? (the howto is not precise enough imho) Quote from the howto: Many CPUs are prone to faulty setups of · memory cache setting (write-back is wrong for BogoMips, often reported lower than 5; write-through is ok) · turbo-buttons (should be ON) · BIOS-software emulated fake cache (change it for real cache) · similar cache and clock related things. Marcus -- Rhubarb is no Egyptian god.Debian GNU/Linuxfinger brinkmd@ Marcus Brinkmann http://www.debian.orgmaster.debian.org [EMAIL PROTECTED]for public PGP Key http://homepage.ruhr-uni-bochum.de/Marcus.Brinkmann/ PGP Key ID 36E7CD09
Re: AMD K6-2 / Bogomips problem
[EMAIL PROTECTED] wrote: besides the fact it seems outdated (1997-12-13), it says like me: AMD K5/K6 clock * (2.00 plusminus 0.010) 11.1 and it lists K6 at 166Mhz having already ~330Bogomips ! so my AMD K6-2 at 350Mhz should have 700Bogomips isn't it ? so next part of my question: what should i check ? bios settings ? (the howto is not precise enough imho) I think I would be concerned too. I have an AMD K6-2 3D 350 MHz processor, overclocked to 400MHz and I get the following from cat /proc/cpuinfo: processor : 0 cpu : 586 model : AMD-K6(tm) 3D processor vendor_id : AuthenticAMD stepping: A fdiv_bug: no hlt_bug : no f00f_bug: no fpu : yes fpu_exception : yes cpuid : yes wp : yes flags : fpu vme de pse tsc msr mce cx8 syscr mmx 3dnow bogomips: 801.18 Sorry if I state the obvious, but you are running at 3.5 x 100 MHz? You do have 100 MHz memory? I have found in the past on a couple of occasions I have had ludicrously low bogomips figures and have re-compiled the kernel (not even changing the settings!) and the rate has gone up the next time i have re-booted. Only ever happened with a RedHat system - don't know why. What kernel are you using? There are a couple of AMD K6-2 patches around. -- Phillip Deackes [EMAIL PROTECTED] Debian Linux v.2.0
Re: AMD K6-2 / Bogomips problem
On 17-Jan-99, Marcus Brinkmann took time to write : On Sun, Jan 17, 1999 at 03:34:42PM +0100, [EMAIL PROTECTED] wrote: besides the fact it seems outdated (1997-12-13), it says like me: AMD K5/K6 clock * (2.00 plusminus 0.010) 11.1 and it lists K6 at 166Mhz having already ~330Bogomips ! so my AMD K6-2 at 350Mhz should have 700Bogomips isn't it ? I don't know what -2 means at K6-2. it's the new generation of K6 chips. also known as K6-3D so next part of my question: what should i check ? bios settings ? (the howto is not precise enough imho) Quote from the howto: Many CPUs are prone to faulty setups of · memory cache setting (write-back is wrong for BogoMips, often reported lower than 5; write-through is ok) less than clear to me · turbo-buttons (should be ON) in these days i rarely found units with turbo buttons anymore · BIOS-software emulated fake cache (change it for real cache) · similar cache and clock related things. yeah... but 'similar' doesn't tell me where to start searching... Thanks for your help, but besides the bios to check - and i'm not very optimistic with that, will see tomorrow - , i don't see anything to do... bad ! Patrick
Re: AMD K6-2 / Bogomips problem
On 17-Jan-99, Phillip Deackes took time to write : so my AMD K6-2 at 350Mhz should have 700Bogomips isn't it ? I think I would be concerned too. I have an AMD K6-2 3D 350 MHz I am ! processor, overclocked to 400MHz and I get the following from cat /proc/cpuinfo: processor : 0 cpu : 586 model : AMD-K6(tm) 3D processor vendor_id : AuthenticAMD stepping: A fdiv_bug: no hlt_bug : no f00f_bug: no fpu : yes fpu_exception : yes cpuid : yes wp : yes flags : fpu vme de pse tsc msr mce cx8 syscr mmx 3dnow bogomips: 801.18 i have : processor : 0 cpu : 586 model : AMD-K6(tm) 3D processor vendor_id : AuthenticAMD stepping: M difference !!! what is it ? fdiv_bug: no hlt_bug : no f00f_bug: no fpu : yes fpu_exception : yes cpuid : yes wp : yes flags : fpu vme de pse tsc msr mce cx8 syscr pge mmx 3dnow ^^^ difference ! what's that ? bogomips: 350.62 i will try to read the kernel sources if i understand anything... but i someone knows about these things... Sorry if I state the obvious, but you are running at 3.5 x 100 MHz? You ^ i have to check tomorrow directly on the main board (is that what you think : multiplier and bus frequency) do have 100 MHz memory? i have two 64Mo chips 100MHz coming from different manufacturers (could that be the cause ?) will bogomips change if memory changes ? I have found in the past on a couple of occasions I have had ludicrously low bogomips figures and have re-compiled the kernel (not even changing the settings!) and the rate has gone up the next time i have re-booted. Only ever happened with a RedHat system - don't know why. isn't that strange ? bogomips should only depend on hardware no ? What kernel are you using? There are a couple of AMD K6-2 patches around. 2.0.36 without any patch i can think of. otherwise the 'compile as [386/486/Pentium/PentiumPro]' option in kernel compilation could change the bogomips rating or not ? i compiled it as a 686. should i compile it as a 586 ? Patrick
Re: AMD K6-2 / Bogomips problem
On Sun, 17 Jan 1999 [EMAIL PROTECTED] wrote: otherwise the 'compile as [386/486/Pentium/PentiumPro]' option in kernel compilation could change the bogomips rating or not ? i compiled it as a 686. should i compile it as a 586 ? From Configure.help: - Pentium for the AMD K5, K6 and K6-3D, Cyrix MediaGX, Cyrix/IBM/National Semiconductor 6x86 and GXm, IDT Centaur WinChip C6, and Intel Pentium/Pentium MMX - PPro for the Cyrix/IBM/National Semiconductor 6x86MX, MII and Intel Pentium II/Pentium Pro I don't know if it would affect bogomips, but it might be worth a try. Bob Nielsen Internet: [EMAIL PROTECTED] Tucson, AZ AMPRnet: [EMAIL PROTECTED] DM42nh http://www.primenet.com/~nielsen