Re: Webalizer weirdness
On Mon, 2007-07-30 at 20:16 +0100, andrew holway wrote: > Hi, > > I've been asking in IRC and forums for some time but no one has been > able to help. > > Has anyone any ideas why this? > http://www.moonet.co.uk/webalizer/usage_200707.html#TOPSITES > > log is here http://www.moonet.co.uk/logs/access.log > and config is here http://www.moonet.co.uk/webalizer/webalizer.conf > webalizer is Webalizer V2.01-10 (Linux 2.6.18-028stab031.dev-ovz031) English > floating point conversion error? > > Linux daisy.moognu.co.uk 2.6.18-028stab031.dev-ovz031 #2 SMP Wed May > 30 10:45:50 BST 2007 sparc64 GNU/Linux Were all of the input data files generated on sparc64? I have seen strange things happen when parsing floating point numbers written on x86 machines (as ASCII). I always assumed it was something to do with the x86 support for 80 bit long doubles (and the sometimes iffy support for 64 bit doubles) verses the SPARC support of 64 bit and (nominally) 128 bit doubles. HTH, Cheers, - Martin -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: NPTL and Debian-SPARC
Aurelien Jarno wrote: You can use the glibc from lenny/unstable. It now uses NPTL on most architectures. Thanks, I'll investigate tom^H^H^H later. -- Mark Morgan Lloyd markMLl .AT. telemetry.co .DOT. uk [Opinions above are the author's, not those of his employers or colleagues] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Sparc32 systems and power consumption
Ludovic Courtès a écrit : > Hi, > > "Bruce O'Neel" <[EMAIL PROTECTED]> writes: > >> With the note that this is getting quite far from debian-sparc's purpose in >> life :-) maybe after this note the conversation should move to the netbsd >> sparc or openbsd sparc lists. > > Maybe it's time to somehow re-spawn Debian GNU/kNetBSD since NetBSD is > apparently paying more attention to avoiding regressions than Linux > (that's a euphemism---but I don't want this to sound too harsh, though, > being myself a humble luser). > We have difficulties to find one or two Linux kernel developers to support the sparc32 port. Do you think it will be easier to find 10 or more developers to do the GNU/kNetBSD sparc port? Reality check please! -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Retiring the sparc32 port
Michelle Konzack a écrit : > Am 2007-07-24 16:16:18, schrieb Aurelien Jarno: >> Michelle Konzack a écrit : >>> And then, there is definitivly a problem to get 32Bit Machines in >>> Strasbourg. All Computer-Stores aelling only those 64Bit CPU's. >>> >> Do you now those 64-bit CPU's are also able to run 32-bit OS? > > But WHA#Y should I pay 100 Euro for a CPU if a 32Bit Sempron with > 2400MHz wil do the job twice? And, Mainboards for 64Bit CPUs are > three times more expensive I advise you to go to another computer store, as this one is clearly swindling you. You can get a boxed Sempron 3000+ in France for less than 35 euros... And a mainboard for less than 55 euros, so if it is three times more expensive, I wonder when you were able to find brand new mainboards for 32-bit CPU for lower than 20 euros... -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: NPTL and Debian-SPARC
Mark Morgan Lloyd a écrit : > I'm trying to track down problems in a development tool (Free Pascal) which > I'm running on Etch. I notice that if I do getconf GNU_LIBPTHREAD_VERSION > it > reports linuxthreads-0.10, while on an x86 Sarge+2.6 system the same command > reports NPTL 0.60. > > Am I correct in believing that Debian for SPARC hasn't yet got NPTL? Is there > any feasible way of putting together a system that /does/ have this > combination, even if it's not of production quality? You can use the glibc from lenny/unstable. It now uses NPTL on most architectures. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Webalizer weirdness
Hi, I've been asking in IRC and forums for some time but no one has been able to help. Has anyone any ideas why this? http://www.moonet.co.uk/webalizer/usage_200707.html#TOPSITES log is here http://www.moonet.co.uk/logs/access.log and config is here http://www.moonet.co.uk/webalizer/webalizer.conf webalizer is Webalizer V2.01-10 (Linux 2.6.18-028stab031.dev-ovz031) English floating point conversion error? Linux daisy.moognu.co.uk 2.6.18-028stab031.dev-ovz031 #2 SMP Wed May 30 10:45:50 BST 2007 sparc64 GNU/Linux Cheers, Andy -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Xorg/ATI failure with Linux 2.6.22 (sparc64)
* Emanuele Rocca <[EMAIL PROTECTED]>, [2007-07-30 17:29 +0200]: > Maybe you can find useful info in the following thread: > http://www.mail-archive.com/[EMAIL PROTECTED]/msg01094.html More precisely, this patch should fix the problem: http://www.mail-archive.com/[EMAIL PROTECTED]/msg01496.html Please keep us informed. :) ciao, ema -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Xorg/ATI failure with Linux 2.6.22 (sparc64)
Hello Ludovic, * Ludovic Courtès <[EMAIL PROTECTED]>, [2007-07-30 17:12 +0200]: > Emanuele Rocca <[EMAIL PROTECTED]> writes: > > Sounds similar to #422077... > > Oops, indeed. > > However, it was closed in May and supposedly fixed in > `xserver-xorg-core_1.3.0.0.dfsg-5', which is apparently not the case (or > it broke again in the meantime?). Yeah, it was just similar. :) Maybe you can find useful info in the following thread: http://www.mail-archive.com/[EMAIL PROTECTED]/msg01094.html ciao, ema -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Xorg/ATI failure with Linux 2.6.22 (sparc64)
Hello Ludovic, * Ludovic Courtès <[EMAIL PROTECTED]>, [2007-07-30 15:55 +0200]: >Fatal server error: >xf86MapPciMem: Could not mmap PCI memory > [base=0x10,hostbase=0x10,size=2000] (Invalid argument) Sounds similar to #422077... ciao, ema -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Xorg/ATI failure with Linux 2.6.22 (sparc64)
Hi, Emanuele Rocca <[EMAIL PROTECTED]> writes: > Hello Ludovic, > > * Ludovic Courtès <[EMAIL PROTECTED]>, [2007-07-30 15:55 +0200]: >>Fatal server error: >>xf86MapPciMem: Could not mmap PCI memory >> [base=0x10,hostbase=0x10,size=2000] (Invalid argument) > > Sounds similar to #422077... Oops, indeed. However, it was closed in May and supposedly fixed in `xserver-xorg-core_1.3.0.0.dfsg-5', which is apparently not the case (or it broke again in the meantime?). Thanks, Ludovic.
Xorg/ATI failure with Linux 2.6.22 (sparc64)
Hi, After upgrading my Ultra 5 to `linux-image-2.6.22-1-sparc64', I noticed that Xorg (ATI driver) would fail to run: X Window System Version 1.3.0 Release Date: 19 April 2007 X Protocol Version 11, Revision 0, Release 1.3 Build Operating System: Linux Debian Current Operating System: Linux sabre 2.6.22-1-sparc64 #1 Tue Jul 24 06:42:28 UTC 2007 sparc64 Build Date: 14 July 2007 Before reporting problems, check http://wiki.x.org to make sure that you have the latest version. [...] (II) Module atimisc: vendor="X.Org Foundation" compiled for 7.1.1, module version = 6.6.3 Module class: X.Org Video Driver ABI class: X.Org Video Driver, version 1.0 (WW) INVALID MEM ALLOCATION b: 0xe100 e: 0xe1ff correcting^G [...] (==) ATI(0): Chipset: "ati". (**) ATI(0): Depth 24, (--) framebuffer bpp 32 (==) ATI(0): Using XAA acceleration architecture Fatal server error: xf86MapPciMem: Could not mmap PCI memory [base=0x10,hostbase=0x10,size=2000] (Invalid argument) >From `strace', the failure looks like this: open("/proc/bus/pci/:01/02.0", O_RDWR) = 7 ioctl(7, IIOCNETDIF, 0) = 0 ioctl(7, IIOCNETSCF, 0) = 0 mmap(NULL, 8192, PROT_READ|PROT_WRITE, MAP_SHARED, 7, 0x10) = -1 EINVAL (Invalid argument) write(2, "\nFatal server error:\n", 21) = 21 The same server works fine with `linux-image-2.6.18-1-sparc64'. The exact version of the Xorg packages are: xserver-xorg 7.2-5 xserver-xorg-core1.3.0.0.dfsg-11 xserver-xorg-video-ati 6.6.3-2 Any hints? Thanks, Ludovic. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
NPTL and Debian-SPARC
I'm trying to track down problems in a development tool (Free Pascal) which I'm running on Etch. I notice that if I do getconf GNU_LIBPTHREAD_VERSION it reports linuxthreads-0.10, while on an x86 Sarge+2.6 system the same command reports NPTL 0.60. Am I correct in believing that Debian for SPARC hasn't yet got NPTL? Is there any feasible way of putting together a system that /does/ have this combination, even if it's not of production quality? -- Mark Morgan Lloyd markMLl .AT. telemetry.co .DOT. uk [Opinions above are the author's, not those of his employers or colleagues] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Debian GNU/(k)NetBSD and sparc32 hardware?
Hi, Ulrich Teichert <[EMAIL PROTECTED]> writes: > What we've seen here is classic bitrot, IMHO. Of course, the main Linux > development platform is x86 and quite a lot kernel developers only work > on one platform. This has introduced bugs for all other ports (and will > continue to do so), which I can understand. Just look at the amount of > patches between 2.6.21 and 2.6.22. Sure, a huge amount of work is being done in between versions, but that new "stable" releases introduce such significant regressions strikes me as a questionable release policy. Of course, developing an OS kernel is a hard task, especially when so many architectures have to be supported, but still. Anyway, I just discovered the Linux Test Project: http://ltp.sourceforge.net/ I guess we, users of those "non economically valuable" architectures, should commit to run LTP every once in a while on latest kernels and report any problems. That might be an improvement given that kernel developers do not seem to run it. > I'm not following NetBSD so closely, so please correct me when something > I write isn't true, but I am under the impression that NetBSD has not > got that much devoted kernel hackers as Linux. As a result, the process > of bitrotting is slower with NetBSD. But NetBSD has a totally different > approach to ports as Linux, just because the motives behind NetBSD are > different. And maybe these reasons will suite [EMAIL PROTECTED] better, > I don't know. Disclaimer: I'm not familiar with NetBSD either, I've never used it actually. I just quickly browsed the web site and mailing list archives, which gave me the impression that when they claim that platform X is supported, it is indeed supported. Nevertheless, you might be right in that bitrotting is just slower on NetBSD than on Linux, it's hard to tell. Thanks, Ludovic. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Retiring the sparc32 port
Michelle Konzack wrote: On the other side, I will leafe in short France and then i will have NO access to Electricity except a bunch (~42) of photovoltaik panels of 75W and a 4kW Bio-Fuel-Generator A 1500Mhz Via C7 (x86 32 bits), a GiB of RAM, a pair of 2.5" hard drives and a slim DVD-burner will happily work with less than 60W, and will run circles around UltraSPARC I and II cpus, let alone SuperSPARC. And will allow for USB, IEEE1394, and other recent stuff to be plugged in. (Anyone know of a VME USB card for my 4/330 ? :-) As an added bonus, AES encryption (ssl, ssh, etc.) will come for almost free, it's implemented in hardware. If you want HW RAID1 instead of SW, an old 3ware PCI RAID1 card for PATA should still fit inside 80W, 100W top. BTW, maybe this part of the discussion should be moved out of debian-sparc now ? It has strayed far from the topic for the list. -- Romain Dolbeau <[EMAIL PROTECTED]> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Retiring the sparc32 port
Am 2007-07-24 16:28:38, schrieb Aurelien Jarno: > Michelle Konzack a écrit : > > My three Sun Blade (32 CPUs, 64 GByte of memory, 96 HDDs) are consuming > > over 4 kW/hour and they are located in Paris/France, Offenburg/Germany > > and Basel/Swiss. > > > > And those are located in the 5% of rich countries that are using far > more energy than the 95% of the others. Righthm since in France I pay for my Dual STM-4 27.000 Euro/month and for electricity 14 ¢/kWh. In germany for a Dual E3 5400 Euro/month and 23 ¢/kWh If I switch to Turkey (my father comes from here) or Iran (my mothers country), I have to pay for a SINGEL E3 (34 MBit) nearly 60.000 Euro and for electricity over 28 ¢/kWh. Switching to the USNA will drop the price for the Backbones but increase the electricity bill dramaticaly... I have already 4 Servers in California and there I pay more for electricity as a whole photopholtaic system cost me... -- American Energy-Politic!!! Thanks, Greetings and nice Day Michelle Konzack Systemadministrator Tamay Dogan Network Debian GNU/Linux Consultant -- Linux-User #280138 with the Linux Counter, http://counter.li.org/ # Debian GNU/Linux Consultant # Michelle Konzack Apt. 917 ICQ #328449886 50, rue de Soultz MSN LinuxMichi 0033/6/6192519367100 Strasbourg/France IRC #Debian (irc.icq.com) signature.pgp Description: Digital signature
Re: Retiring the sparc32 port
Am 2007-07-24 16:16:18, schrieb Aurelien Jarno: > Michelle Konzack a écrit : > > And then, there is definitivly a problem to get 32Bit Machines in > > Strasbourg. All Computer-Stores aelling only those 64Bit CPU's. > > > > Do you now those 64-bit CPU's are also able to run 32-bit OS? But WHA#Y should I pay 100 Euro for a CPU if a 32Bit Sempron with 2400MHz wil do the job twice? And, Mainboards for 64Bit CPUs are three times more expensive So WHY ahould I buy Oversized mainboard+64Bit-cpu+memory+Graphiccard for over 200 Euro IF I do not need it? AND, since I need my computer 24/7 and CAN NOT SWITCH into Power-Save mode I wo not want to pay per day 0.82 Euro electricity for it. Please note, that I have more then one Server/Workstation @home and I pay already for my computers 280 Euro elektricity per month. On the other side, I will leafe in short France and then i will have NO access to Electricity except a bunch (~42) of photovoltaik panels of 75W and a 4kW Bio-Fuel-Generator Currently I am switching my Workststions to arm9 but having problems getting external Graphiccards or SBC's based on arm/mips with graphic on board. Thanks, Greetings and nice Day Michelle Konzack Systemadministrator Tamay Dogan Network Debian GNU/Linux Consultant -- Linux-User #280138 with the Linux Counter, http://counter.li.org/ # Debian GNU/Linux Consultant # Michelle Konzack Apt. 917 ICQ #328449886 50, rue de Soultz MSN LinuxMichi 0033/6/6192519367100 Strasbourg/France IRC #Debian (irc.icq.com) signature.pgp Description: Digital signature
Re: Debian GNU/(k)NetBSD and sparc32 hardware?
Hi Ludovic, I wrote: >> That's not quite true. Dave Miller is still collecting patches, Mark >> Fortescue, Krzysztof Helt and others are producing them. See the respective >> posts on [EMAIL PROTECTED] It's just that there is no real maintainer >> for the port. > >[...] > >> Call me a chicken, but I still think it will be less work to just fix the >> issues in the kernel and use the existing stuff instead. > >That's the whole issue. I am under the impression (perhaps wrongfully) >that Linux development is moving at a high pace, not considering support >of "legacy" hardware as a high priority. For instance, the first 2.6 >releases introduced significant regressions wrt. SPARC32 support >compared to 2.4. Only now is 2.6 starting to catch up with 2.4, thanks >to the work of a few people. What we've seen here is classic bitrot, IMHO. Of course, the main Linux development platform is x86 and quite a lot kernel developers only work on one platform. This has introduced bugs for all other ports (and will continue to do so), which I can understand. Just look at the amount of patches between 2.6.21 and 2.6.22. >Conversely, it seems that NetBSD values continued support more, judging >from the mailing list archives of various ports (including, e.g., the >still active VAX port!). It's probably following a much more >conservative development approach, less biased towards newer hardware. I'm not following NetBSD so closely, so please correct me when something I write isn't true, but I am under the impression that NetBSD has not got that much devoted kernel hackers as Linux. As a result, the process of bitrotting is slower with NetBSD. But NetBSD has a totally different approach to ports as Linux, just because the motives behind NetBSD are different. And maybe these reasons will suite [EMAIL PROTECTED] better, I don't know. >> I agree that a new debian architecture would be more fun, but >> splitting up the remaining debian sparc32 developers between NetBSD >> and Linux does not sound too healthy for me. > >Agreed. In the short term, it does seem "easier" to try and fix Linux' >SPARC32 support. However, I'm wondering whether that would be a good >long-term investment. > >Now, similar issues may also arise with other architectures, too. Right. m68k and mipsel are springing to my mind. Well, as I am not a debian developer, I probably should keep my mouth shut now and leave the discussion (and of course the final decision) for long term or short term investment to debian members which care about sparc32. CU, Uli -- Dipl. Inf. Ulrich Teichert|e-mail: [EMAIL PROTECTED] Stormweg 24 |listening to: Channel 13 Is Haunted (Hex Dispensers) 24539 Neumuenster, Germany|Is It Good Or Is It Bad? (Opération S) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Help with Linux on sparc64 (Ultra 45)
Hi. I have a Sun Ultra 45 (hardware works, Solaris loads fine) and I've tried (unsuccessfully) to install some kind of Linux distribution on it. I've tried Debian netinst (both stable and testing), but after boot the keyboard stops working (It only has USB keyboard&mouse). Booting from the serial console, It hangs with the following message: su: Cannot register IRQ 1 I've tried Gentoo (universal), Ubuntu and Aurora but they either have the same problem or fail to boot entirely. The machine can netboot (note: If you want to netboot a Sun machine, use Debian's tftpd and not Fedora Core), and I can reach the 'choose language' menu (but no keyboard input). So it seems to me I can put the hard drive in a normal PC and 'dump' a debian sparc unto it. Is there a manual for this kind of installation? Has anyone had any luck with this particular machine? What binary distribution has the best support for Sun machines? Can anyone help me? Nadav Shemer [EMAIL PROTECTED]
Re: Debian GNU/(k)NetBSD and sparc32 hardware?
Ludovic Courtès wrote: Hi, BERTRAND Joël <[EMAIL PROTECTED]> writes: Maybe, but NetBSD kernel does not correctly work on sun4m/SMP, like Linux. Today, no one OS can be used on sun4m/SMP workstations, and I think that it will be easier to fix linux 2.6 sparc32 kernel than work on debian/xBSD sparc32 port. I had reports stating the contrary, as far as HyperSPARC SMP is concerned (which seems like the less-likely-to-work configuration): http://article.gmane.org/gmane.os.netbsd.ports.sparc/6945 http://article.gmane.org/gmane.os.netbsd.ports.sparc/6943 OTOH, there were mixed reports related to SMP with other processors: http://article.gmane.org/gmane.os.netbsd.ports.sparc/6931 http://article.gmane.org/gmane.os.netbsd.ports.sparc/6914 Ludovic, I've never said that NetBSD does not boot, but that NetBSD is not stable enough to be in production state. In SMP configuration (SS or HS), I've never seen any stable NetBSD kernel (since 3.0) on sparc32. It randomly panics or hangs on all my sparc32/SMP. NetBSD 4.0 kernel has a broken SX fb support, and maybe my tests are not significative. Regards, JKB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Debian GNU/(k)NetBSD and sparc32 hardware?
Ludovic Courtès wrote: Hi, BERTRAND Joël <[EMAIL PROTECTED]> writes: Maybe, but NetBSD kernel does not correctly work on sun4m/SMP, like Linux. Today, no one OS can be used on sun4m/SMP workstations, and I think that it will be easier to fix linux 2.6 sparc32 kernel than work on debian/xBSD sparc32 port. I had reports stating the contrary, as far as HyperSPARC SMP is concerned (which seems like the less-likely-to-work configuration): http://article.gmane.org/gmane.os.netbsd.ports.sparc/6945 http://article.gmane.org/gmane.os.netbsd.ports.sparc/6943 OTOH, there were mixed reports related to SMP with other processors: http://article.gmane.org/gmane.os.netbsd.ports.sparc/6931 http://article.gmane.org/gmane.os.netbsd.ports.sparc/6914 Ludovic, I've never said that NetBSD does not boot, but that NetBSD is not stable enough to be in production state. In SMP configuration (SS or HS), I've never seen any stable NetBSD kernel (since 3.0) on sparc32. It randomly panics or hangs on all my sparc32/SMP. NetBSD 4.0 kernel has a broken SX fb support, and maybe my tests are not significative. Regards, JKB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]