Re: Which Sparc is best?
Hello On Tue, 23 Jul 2002, Joshua Uziel wrote: The U5 and U10 are decent machines... moreso on the systems with the 2MB ecache (most of the U10s, and some of the U5s). One problem with those systems is that they have a CMD646U for an EIDE controller... a buggy chipset that Linux deals with poorly (mostly due to CMD's lack of desire to cooperate, I hear). On Wed, 24 Jul 2002, Ferris McCormick wrote: For what it's worth, my experience with Linux (Debian SuSE) on an Ultra10 suggests that if you are going to stress the disk much at all, on such a system, you will want to use one of your expansion slots for a SCSI card. Others will have better information, though. On Wed, 24 Jul 2002, Craig Morehouse wrote: This makes much sense. I use SCSI on all the important Intel boxes. FWIW, my Ultra5/IDE box experienced frequent disk troubles even when it wasn't stressed much (desktop). It kept randomly changing characters in files, leading to visible problems say once per 2-3 months[1]. When stressed more (mini-server for 4-5 people), the bug visibility raised to about one per 2-3 weeks, and eventually led to an unusable system. Both under Debian pre-woody and Solaris (I tested Solaris in order to see whether it would improve, but nope -- my impression was that it became even worse). So, I think you would be probably better off with SCSI even for desktops. Tibor Footnotes: [1] Or faster, almost always, when untarring huge files. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: unstability of woody ...
Hello hi, i'm using potato on a ultra5 and i'd like to try woody ... There may not be any problems now (there might be, but I don't know all of the details), but sooner or later shit will hit the fan. E.g. if your Ultra 5 contains atyfb (ATY 3D RAGE PRO) you probably would not be able to run woody's X11 4.0.1 - there were problems with linux-sparc port of ATI driver. (is this still an issue?) cheers -- TS
Re: sparc vs apollo
Hello Does anyone here know how a HP Apollo Model 715 compares with a Sparc. For instance is a Apollo 715 similar in terms of cpu horsepower to a sparc lx, sparc 4, sparc 5, etc? SPEC URL:http://www.spec.org/ is your friend, e.g: model SPECint95 SPECfp95 -- - HP 9000 Series 700 Model 715/100 2.89 3.47 Sun SPARCstation 5 model 170 3.53 3.00 Sun Ultra 5 model 2709.1710.6 ... cheers -- TS
Re: netscape and others ...
Hello I run libc6 2.1.94-3, but I'm of course not completely sure it's glibc that's causeing the bus error, might be one of the other libraries. FWIW I was running an older version of libc6 with woody. As soon as I upgraded to libc6 2.1.95-1 the bus error appeared... )-: cheers -- TS
Re: netscape and others ...
Hello When I upgraded my SS5 to latest woody, which means a new glibc, this version of netscape stopped working. I'm using `navigator-smotif-45_4.5-1_sparc.deb' which works well with woody; if I'm not wrong you can get it at ftp://ftp.mech.kth.se/pub/nordmark/debian/dists/slink/non-free/binary-sparc/web/. cheers -- TS
Re: Mozilla (was: Re: getty oops and netscape)
Hello PH == Peter Haworth [EMAIL PROTECTED] writes: PH I don't find that. I get this error: PH Could not obtain CmdLine processing service Likewise, followed by gobs of Javascript errors. I filed a bug report. Does anyone have any suggestions for things to try to find the problem, which I assume is SPARC-specific? It's most probably Debian-SPARC-specific only, since precompiled M17 linux-sparc binary tarball from mozilla.org works fine. Check it out. cheers -- TS
Re: Keyboard map under X
Hello the keys seemed to be mis-mapped. Can anyone help me fix this? apt-get install xkeycaps cheers -- TS
Re: test this ! (hard lockups)
Sergey V Kovalyov [EMAIL PROTECTED] writes: strace -ff -p `pidof ident` and telnet from another machine to e.g. port 113 Assuming you meant inetd here. My Ultra-5 box (synced with current potato) survives without problems. cheers -- TS
Re: Ultra 5 reproducible XF86_Mach64 lockups
Hein Roehrig [EMAIL PROTECTED] writes: Just FYI, I can reproduce the problem here as well (almost same setup, but stock 2.2.15). A good news to share: all my lockup problems are apparently gone after upgrading to the latest potato! (namely, xserver-mach64 to 3.3.6-7 and kernel-image-2.2.15-sun4u to 2.2.15-2) cheers -- TS
Ultra 5 reproducible XF86_Mach64 lockups
Hello I obtain repetitive lockups of Ultra 5 under X11, Sun Type 5 keyboard; kernel 2.2.15, synced with current potato. The lockup typically occurs within 1 minute of intensive X input. To reproduce it I use xdvi and emacs and rapidly click with mouse here and there, change scales, skip through pages, reread file, etc. The lockup happens independently of the window manager (tested E, sawmill, fvwm). When the lockup happens both the mouse and keyboard are unusable. Machine respond to network connections and top shows that XF86_Mach64 is working hard. But I have not succeeded to kill X, nor to reboot over the network connection: after such attempts the connection itself is frozen. The only apparent solution is to cut the power off. I wonder whether anyone observes the same thing. What is the best method for further diagnosis? cheers -- TS
Re: Ultra 5 reproducible XF86_Mach64 lockups
Ben Collins [EMAIL PROTECTED] writes: Are you using the 2.2.15 kernel from potato, or one you compiled yourself from stock 2.2.15 source? The native potato one (2.2.15-0.19.4, from kernel-image-2.2.15-sun4u). cheers -- TS
Re: Ultra 5 reproducible XF86_Mach64 lockups
Ben Collins [EMAIL PROTECTED] writes: Interesting. That source is working fine here on a U5 and a U10 class system. What all are you running under X? What, if anything, are you running on the system besides X related programs? The system was basically idle (nothing special apart from X server). Moreover, in order to diagnose the problem I've tried to run as little stuff as possible. The freezing happened with: fvwm + gnome-terminal (only terminal; not the full GNOME) + emacs + xdvi. I haven't tried to narrow it further. BTW I've seen also a minor unreproducible display problem with current XF86_Mach64 (trash displayed here and there on the screen; no lockup). Perhaps it would be interesting to build the new X11 and test it. Hein Roehrig [EMAIL PROTECTED] writes: Just FYI, I can reproduce the problem here as well (almost same setup, but stock 2.2.15). Now this gets really interesting! :) cheers -- TS