Re: gpm does not work
| 1) gpm does not work. Whe i run "gpm -t sun -m /dev/rtc" it does | nothing, even return the call. /dev/rtc? Are you kidding? This is the real time clock device, certainly not appropriate for a mouse! :) | 2) The X servers don't work, I have tried all of them. Moreover, I | don't know how to configure it as xf86config and XF86Setup are not | present. Did you try /dev/rtc as the frame buffer device? Just kidding :)
Re: New modutils uploaded
>>>>> "Wichert" == Wichert Akkerman <[EMAIL PROTECTED]> writes: Wichert> I just uploaded motudils 2.3.7-1. This should fix all the Wichert> problems with sparc32/sparc64 systems.. can somone confirm Wichert> that? Grmpf, I'd love to answer, but: gcc -O2 -Wall -I../../../insmod/../include -D_GNU_SOURCE -DCOMPAT_2_0=1 -DCOMBINE_sparc -DELF_MACHINE_H='"elf_sparc.h"' -DARCH_sparc -DCOMBINE_rmmod -DCOMBINE_modprobe -c -o insmod.o ../../../insmod/insmod.c ../../../insmod/insmod.c: In function `print_load_map': ../../../insmod/insmod.c:343: internal error--unrecognizable insn: (insn 704 703 6 (unspec_volatile[ (mem (mem:SI (plus:SI (reg:SI 272) (const_int 8 [0x8])) 0) 0) ] 4) -1 (nil) (expr_list:REG_DEAD (reg:SI 272) (nil))) make[2]: *** [insmod.o] Error 1 (gcc 1:2.95.2-3, which seems to be the latest version) Sam -- Samuel Tardieu -- [EMAIL PROTECTED]
Re: call for help with failed sparc package builds
As far as Python is concerned, the autobuilder tries to use tcl8.2-dev and tk8.2-dev while Python expects tcl and tk 8.0. How does the autobuilder choose what development package to use? I've uploaded a new version of python and friends, and asked the maintainer to switch to 8.2 if it works fine. In the meantime, there is a definite glitch in the autobuilder. pgpPJTDTSELzE.pgp Description: PGP signature
Re: call for help with failed sparc package builds
Ok. I just uploaded a2ps_4.12c-1_sparc.deb to master. I cannot understand how your autobuilder got this error: Making install in emacs make[3]: Entering directory `/scratch/buildd/a2ps-4.12c/contrib/emacs' make[4]: Entering directory `/scratch/buildd/a2ps-4.12c/contrib/emacs' make[4]: Nothing to be done for `install-exec-am'. /bin/sh ../../auxdir/mkinstalldirs /usr/bin/install -c -m 644 make-regexp.el /make-regexp.el /usr/bin/install: cannot create regular file `/make-regexp.el': Permission denied /usr/bin/install -c -m 644 a2ps.el /a2ps.el /usr/bin/install: cannot create regular file `/a2ps.el': Permission denied /usr/bin/install -c -m 644 a2ps-print.el /a2ps-print.el /usr/bin/install: cannot create regular file `/a2ps-print.el': Permission denied make[4]: *** [install-lispDATA] Error 1 That would mean that $(prefix) is wrong here (I cannot believe that automake has such a failure). $(prefix) seems to be void, and I cannot explain why. pgpUpyWvfM0Wy.pgp Description: PGP signature
Re: Where is sparc64-linux-ld?
On 7/09, Eric Delaunay wrote: | PS: I have had troubles compiling VORTEX (3c59x) & QLOGIC ISP support (egcs64 | segfault). I needed to disable them. I use a "special" patch around those two files: #if 0 ... #endif :)
Re: Freeze of 2.2.9/2.2.10 on Ultra5
>>>>> "Ben" == Ben Collins <[EMAIL PROTECTED]> writes: Ben> There was a version of egcs64 in potato that did cause problems Ben> like you describe. The current egcs64 is known to be good. New crash yesterday and today at 4AM. After investigation, it looks like the cluprit is this crontab: 0 4 * * * postgres [ -x /usr/lib/postgresql/bin/do.maintenance ] && /usr/lib/postgresql/bin/do.maintenance -a I just run it by hand and I got a full freeze (no three fingers key, no magic sysreq, no L1-A, etc.). Is anyone around running Postgres on a Ultra5 with a recent kernel? Sam -- Samuel Tardieu -- [EMAIL PROTECTED]
Re: Freeze of 2.2.9/2.2.10 on Ultra5
On 26/08, Ben Collins wrote: | What version of egcs64 are you using to compiled the kernel? I suggest | making sure you use the latest one in potato right now. I was using the egcs64 from potato at the time of the compilation. I just recompiled 2.2.12 with the current egcs64, and I have had no crash so far, but the machine has only been up for two hours. If it crashes again during the next few days, I'll drop a note here.
Re: Boot question
On 26/08, [EMAIL PROTECTED] wrote: | STOP-A, and at the OK prompt type: | | setenv boot-device disk:a Thanks, works perfectly.
Boot question
Sorry if this is a FAQ, I couldn't find the answer. My Ultra5 tries to boot with "boot disk:b", and my Linux partition is disk:a (hda1). How do I change this in the PROM? Or how do I install a boot block at the beginning of hda2 which is the swap V1?
Freeze of 2.2.9/2.2.10 on Ultra5
I get full freezes regularly with kernels 2.2.9 (Debian version) and 2.2.10 (compiled locally) on an Ultra5. Sometimes the machine runs fine for two days, sometimes it crashes again after one hour. I cannot even Stop-A to get in the monitor, I have to go through a full power cycle. Any hint? Sam -- Samuel Tardieu -- [EMAIL PROTECTED]
Where is sparc64-linux-ld?
When recompiling a kernel with "make-kpkg --arch sparc64 ..." on an Ultra5: sparc64-linux-ld -r -o kernel.o signal.o ksyms.o sched.o dma.o fork.o exec_domain.o panic.o printk.o sys.o module.o exit.o itimer.o info.o time.o softirq.o resource.o sysctl.o acct.o capability.o kmod.o make[3]: sparc64-linux-ld: Command not found Where can I find it? Sam -- Samuel Tardieu -- [EMAIL PROTECTED]
Re: Mod_perl and Linux/Sparc
On 13/08, Ben Collins wrote: | Can you send me a list of what you had previously? Not very useful: everything turned on :))) (with all optional modules)
Re: Mod_perl and Linux/Sparc
On 13/08, Ben Collins wrote: | Here is the complete list of the modules I have loaded: Grmpf, it works fine. Now, we have to find which combination causes those trouble...
Re: Mod_perl and Linux/Sparc
>>>>> "Ben" == Ben Collins <[EMAIL PROTECTED]> writes: Ben> Then perhaps it was a mixture of loaded/unloaded modules Ben> conflicting. I had been fighting with it for about 4 days, when I Ben> finally uncommented one of the auth modules and it worked. I had Ben> been so fed up with it, that I didn't bother changing anything Ben> else to pinpoint it (it worked and I wasn't going to mess with it Ben> anymore :) I'd be interested in a "grep LoadModule /etc/apache/httpd.conf", so that I can try your very configuration. Sam -- Samuel Tardieu -- [EMAIL PROTECTED]
Re: Mod_perl and Linux/Sparc
On 13/08, Ben Collins wrote:
| Nasty, that actually looks like it's in libc, -lresolv or nss_files. I
| don't think apache nor php3 would access /etc/hosts directly.
Yup, that's right. Just before that, we have:
connect(3, {sun_family=AF_UNIX, sun_path="/var/run/.nscd_socket"}, 110) = -1
ECONNREFUSED (Connection refused)
and apache does not know about nscd.
Re: Mod_perl and Linux/Sparc
>>>>> "Sam" == Samuel Tardieu <[EMAIL PROTECTED]> writes: Sam> Thanks! Same think for me, it now starts. So I would guess that Sam> the bug is not in the php3 module itself, but in one of those... Argh, forget it: I have apache processes now, but nothing on Port 80. -- Samuel Tardieu -- [EMAIL PROTECTED]
Re: Mod_perl and Linux/Sparc
>>>>> "Gergely" == Gergely Madarasz <[EMAIL PROTECTED]> writes: Gergely> hmpf... thunderchild:~# grep auth /etc/apache/httpd.conf Gergely> #LoadModule auth_module /usr/lib/apache/1.3/mod_auth.so Gergely> #LoadModule anon_auth_module Gergely> #/usr/lib/apache/1.3/mod_auth_anon.so LoadModule Gergely> #dbm_auth_module /usr/lib/apache/1.3/mod_auth_dbm.so Gergely> #LoadModule db_auth_module /usr/lib/apache/1.3/mod_auth_db.so Gergely> and it starts... Thanks! Same think for me, it now starts. So I would guess that the bug is not in the php3 module itself, but in one of those... -- Samuel Tardieu -- [EMAIL PROTECTED]
Re: Mod_perl and Linux/Sparc
>>>>> "Ben" == Ben Collins <[EMAIL PROTECTED]> writes:
Ben> Personal experience. It requires one of the auth modules in
Ben> apache to be loaded, or php3 wont load itself. I don't know how
Ben> you can check for this reliably in the pre/post scripts
Ben> though. I'm not even sure now which one of the modules it was (I
Ben> know it was auth related though).
Looks like this is related to /etc/hosts. Too bad, we do not have
ltrace on Sparc...
open("/etc/hosts", O_RDONLY)= 3
fcntl(3, F_GETFD) = 0
fcntl(3, F_SETFD, FD_CLOEXEC) = 0
fstat(3, {st_mode=0, st_size=934587505, ...}) = 0
mmap(0, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) =
0x7001c000
read(3, "127.0.0.1\tlocalhost\n137.194.16"..., 8192) = 306
--- SIGILL (Illegal instruction) ---
detach: ptrace(PTRACE_DETACH, ...): Input/output error
--
Samuel Tardieu -- [EMAIL PROTECTED]
Re: Mod_perl and Linux/Sparc
On 13/08, Ben Collins wrote: | It's already tested and works. I just have to do it is all, but thanks any | way :) Ok :) Btw, I just used the nice "dpkg-repack libgmp2" on a machine with a working ssh and installed the recreated package on the Ultra5, and things work fine.
Re: Mod_perl and Linux/Sparc
On 13/08, Ben Collins wrote: | There isn't a working mainstream 64bit environement yet. The reason ssh | doesn't work is because of a broken libgmp2 currently in the sparc/potato | dist. This is due to the patch in the previous version (fixing this | problme once before) not being incorporated by the libgmp2 maintainer. I | thought some one had already done and NMU, but since it hasn't been done, | I'll add it to my todo list for this weekend. If you send me the patch (if you have it), I can test it right away and do a NMU of libgmp2 if ssh works well.
Re: Mod_perl and Linux/Sparc
On 13/08, Ben Collins wrote: | I think that might be attributed to the dependency on another module (I | think it requires one of the auth modules, forget which one), or else it | wont start, nor will it give an error message as to why (this is for | php3). Why would it restart then? Another question: where can I find /usr/lib64/* files? I tried to recompile ssh because it does not work at all on my Ultra5, and it still gets me the same error (loop in crypto routines). I wanted to try it in 64 bits mode (maybe it's stupid), but egcs64 does not contain /usr/lib64/crt1.o.
Re: Mod_perl and Linux/Sparc
>>>>> "Adam" == Adam Di Carlo <[EMAIL PROTECTED]> writes: Adam> Ryan, your mod_perl problems sound wierd. Can you file a bug Adam> against mod_perl? Well, it really looks like my mod_php3 problem... I just installed unstable on an Ultra5, and it gives me the same behaviour as on the Sparc10: apache can be *reloaded* (apachectl restart) with mod_php3, but cannot start with it. Sam -- Samuel Tardieu -- [EMAIL PROTECTED]
Re: glibc2.1 and potato
On 12/05, Ben Collins wrote: | All the new packages will run with slink's glibc, they are perfectly | compatible. Please report any problems to the contrary as we are set on | having this compatability. Btw, I am running glibc2.1.1-2 on my Sparc10 (Sun4m) without any apparent problem (as long as I don't try to use nscd). I have a 2.2.7 with the "break" patch.
Re: Fwd: [PATCH] oneliner in iommu.c
On 7/05, Heikki Vatiainen wrote: | Just in case someone is not reading linux-sparc, I'm reposting | this small patch. After this patch my SS20 boots 2.2.7 kernel and | seems to be stable. Thanks, I don't read it. I'm recompiling the kernel with this patch, hopefully I'll be able to use the new glibc2.1 then :)
Re: [NOTICE] libc6 problems (Re: libc and other things)
>>>>> "Juziel" == juziel <[EMAIL PROTECTED]> writes: Juziel> Eh? I tried building and running my own 2.2.7 both stock and Juziel> from CVS on my SS10... and under both Debian 2.1 and RH 6.0 Juziel> ... none of those cases would boot 2.2.7 for me. So, the question now is: has anyone be able to boot a SS10 using kernel 2.2.7? Sam -- Samuel Tardieu -- [EMAIL PROTECTED]
Re: [NOTICE] libc6 problems (Re: libc and other things)
On 5/05, Ben Collins wrote: | That's just not possible...been thought of, and there is no real | solution. I suggest for people testing _unstable_ to seriously consider | installing sash (a static compiled shell) and always keep .debs around | of a working libc6. Thanks for the advice, it's now installed. | Hate to say it folks, but this is unstable, and it can and will break at | any moment. So if your not prepared to take measures for fallback, or | you can't afford downtime on the system, stick with stable :) I was not complaining :) Although I use SparcLinux on a production machine, I'm willing to play games sometimes to help development. Now that I've recovered from a failure once (it took less than 15 minutes to setup the tftp boot and repair everything), it will even take less time now, so I'm ready to try other libc and kernels.
Re: libc6 2.1.1, kernel 2.2.6 and Sun4m
On 5/05, Ben Collins wrote: | | Nope, it's due to some brokeness that we need to fix pronto, I am | uploading and one today in time for dinstall into the archive. The only | change is sun4m requiring >=2.2.7 kernel, this is until we find a | better solution. 2.2.7? I thought you had the same problem (or maybe it was someone else, but it was posted on the list) as me, i.e. 2.2.7 not booting on Sun4m.
libc and other things
May I suggest that we introduce a package named libc60 that could be left installed to ease downgrading when/if we have problems with the newest libc? Sam -- Samuel Tardieu -- [EMAIL PROTECTED]
Re: libc6 2.1.1, kernel 2.2.6 and Sun4m
On 4/05, Ben Collins wrote: | Was this the glibc 2.1.1-2? If so, it should have been fixed, I tested | it myself on a sun4m/2.2.5(non-cvs) and Steven tested on a | sun4m/2.2.1(stock debian image). If that's the case we need to do some | quick fixing. Yup: glibc (2.1.1-2) unstable; urgency=low I got a big screwup. Could it be due to the fact that I'm running a stock 2.2.6 kernel?
libc6 2.1.1, kernel 2.2.6 and Sun4m
I thought everything was working fine (from what I read here, but I guess I got it wrong) :/ init didn't boot. Nothing worked. I had to boot from the net to get the old libc from other machines. It was horrible :-) Known problem? :)
Re: Apache and php3
On 4/05, Gergely Madarasz wrote: | It starts for me ok, but there is another problem with php3-pgsql: | it seems to be statically linked against libpq. Thanks for trying. I don't understand why it does this for me, I guess apache has been compiled against an old libc while php3 has been compiled against a new libc, which may explain the problem. I'm rebuilding apache with my current libc to see how it behaves.
Re: Apache and php3
On 4/05, Samuel Tardieu wrote: | I'd be interested if someone could try to reproduce my problem: on a Sparc, | install apache and php3 (that's a module), and try to run a server with | this module enabled. | | My apache refuses to start when this module is enabled in httpd.conf, | but will restart if I enable this module after Apache has been started | (apachectl restart). Quite odd... One thing I forgot: php3-pgsql is installed too.
Apache and php3
I'd be interested if someone could try to reproduce my problem: on a Sparc, install apache and php3 (that's a module), and try to run a server with this module enabled. My apache refuses to start when this module is enabled in httpd.conf, but will restart if I enable this module after Apache has been started (apachectl restart). Quite odd...
Sparc10 cannot boot kernel 2.2.7
Uncompression image... PROMLIB: obio_ranges 5 Booting Linux... Then nothing more. Even STOP-A doesn't work, off/on required. I use the package kernel-source-2.2.7. 2.2.6 works fine however, except for PPP (pppd keeps reporting that PPP has not been compiled in the kernel, despites the fact that the PPP module has been succesfully loaded).
Sparc 10 and kernel 2.2.6
-BEGIN PGP SIGNED MESSAGE- FYI, I'm running a vanilla 2.2.6 kernel on a SparcStation 10 and it seems to run fine (WWW server, heavily loaded mail server, PostgreSQL server). Sam - -- Samuel Tardieu -- [EMAIL PROTECTED] -BEGIN PGP SIGNATURE- Version: 2.6.3ia Charset: latin1 iQCVAwUBNx8AcYFdzKExeYBpAQGZ6QP+I5SmLbGOpZOWVlc0rqLnD0xKlZpNN54i Qwt/7JH6yMzndlqcYa0EFRNSjTFBeSPSnt3cxC26qVnoaGfSR7nRDJb5Vrwm+EnQ VnrhUKypl88IH4QnadpkuF+jITEVDnyE3wVD0P1gTx7mb/DY4scWd1X2AxvJ99/V fVnLNBnjhis= =my0X -END PGP SIGNATURE-
Xsun eating a lot of CPU on sun4m
I have two sun4m (Sparc10) running the latest unstable, and Xsun always takes a lot of CPU while the machine only have X, fvwm with no fancy feature and a xterm with nothing happening inside (in fact, one of the machines is even waiting for a xdm login). 4989 root 20 0 7144 7144 1744 R 0 43.8 25.2 0:45 cc1 2128 root 20 0 2372 1628 400 R 0 35.0 5.7 30:37 Xsun 5003 root 20 0 1068 1068 808 R 0 23.8 3.7 0:01 top 4998 root 19 19 872 872 440 R N 0 0.4 3.0 0:00 cpp 1 root 0 0760 0 SW 0 0.0 0.0 0:01 init 2 root 0 0 00 0 SW 0 0.0 0.0 0:05 kflushd 3 root 0 0 00 0 SW 0 0.0 0.0 0:07 kswapd 12 root 0 088 5228 S 0 0.0 0.1 0:07 update 1181 root 0 0 268 184 144 S 0 0.0 0.6 0:09 syslogd Usually (when I have no compilation running), Xsun takes about 90% of the CPU. No screen saver, nothing... Any idea? Sam -- Samuel Tardieu -- [EMAIL PROTECTED]
Re: The road ahead ...
On 27/01, Jules Bean wrote: | > When we say fakeroot is broken, do we mean the one in libtricks, or the | > actual fakeroot package. The libtricks version seems to have a problem | > on glibc 2.1 systems, this was verified by some people using the other | > ports. I just tries the actual fakeroot package and it seemed to work | > fine. | | I mean libtricks. | | Is fakeroot still available for slink? Presumably it is, and I did | something stupid (I did try to find it, but I could have made some foolish | mistake). apt-get install fakeroot (it worked just fine for me 3 hours ago when I noticed the segmentation fault with the libtricks one) Sam -- Samuel Tardieu -- [EMAIL PROTECTED]
Re: eject floppy in SS10?
>>>>> "Matt" == Matt Wagner <[EMAIL PROTECTED]> writes: Matt> i seem to have a bit of a problem. :) how the heck to i get the Matt> floppy out of my SS10 machine... there's got to be a better way Matt> other than the paperclip! :) apt-get install eject ? Sam -- Samuel Tardieu -- [EMAIL PROTECTED]
Re: zsh ouch
On 26/11, Jules Bean wrote: | Did someone just upload a new zsh? | | It segfaults when run on my SS2/2.1.125... You mean "bus error"? I've reported this yesterday as a zsh bug, but didn't receive any feedback yet. Sam -- Samuel Tardieu -- [EMAIL PROTECTED]
