Bug#464752: Bug#491505: [package varmon] varmon segfaults on Etch i386
Am Tue, 26 Aug 2008 08:38:27 +0200 schrieb Alessandro Polverini [EMAIL PROTECTED]: Christoph Franzen wrote: [...] I wonder if you suffer the same problem tough. The problem I suffered was visible only in certain condition, i.e.: when a lan user accesses the samba shares. Locally I saw no big differences apart some performance issues on the startup of tomcat but these are way harder to replicate. So I think the remote access is not really useful if I've not some windoze user that can do some remote (for me, local for you) tests on the server machines. This is possible, installing Samba is no problem, and I can do either some manual tests or set up an automatic job (for example mounting the samba share from my backup server and copying files around), if you tell me how to reproduce the problem. Christoph signature.asc Description: PGP signature
Bug#464752: Bug#491505: [package varmon] varmon segfaults on Etch i386
Am Mon, 25 Aug 2008 11:02:55 +0200 schrieb Alessandro Polverini [EMAIL PROTECTED]: Ciao Allessandro, I see you have a Mylex Acceleraid RAID adaptor, I've got three DAC960 controllers (2 of them dual [or three?] channel SCSI III UW LVD Acceleraid 150 [or was it 170?] aka DAC960PTL1 with 8 MiB EDO RAM, and one older single channel DAC960PL fast wide SE SCSI II with 32 MiB FPM RAM with ECC. The first two are located in a stand by production system of my brother, and in my backup and APT cache server. On these systems only limited tests are possible, and I would have to perform them on my own, because these systems must be fully functional in case the primary systems are not. The older controller, however, resides in a Debian Lenny test system fully connected to the internet, and I can bring it back online at any time for testing purposes. have you problems like me with newer kernels? http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=464752 In fact, I have never worried about performance issues, so I can't say anything about that bug. That I haven't noticed anything of course doesn't mean that I don't have this problem. I would like to do more tests but the problem is that machine is in production and I've no time and capability to better check the problem. I wonder if you suffer the same problem tough. I can switch on the test machine at any time and provide you, the Debian or upstream developers with the connection data via encrypted mail. Simply send me your key fingerprints and mail addresses. You will have full root access and internet connectivity around the clock via ADSL (no flatrate though, so keep me informed of huge data transfers to or from the machine). The machine is not in use, so you can install kernels and reboot at will. Compiling will not be fast (Pentium II 400 MHz, all disks are older 8 GB on the RAID controller), however. Two of the disks are empty, and can be used for performance tests. Regards, Christoph signature.asc Description: PGP signature
Bug#491505: [package varmon] varmon segfaults on Etch i386
Am Sun, 17 Aug 2008 09:52:56 +0200 schrieb Julien Danjou [EMAIL PROTECTED]: I'm preparing a new upstream release and will push it back to Debian ASAP (expect everything tomorrow). Thank you for fixing the bug, it works for me now, no problems so far. Thanks for the box Christoph. It was very useful. You can probably shut it down now. I'd suggest to keep it around just in case of a new bug in the coming weeks if you can, we never know, but that'd be odd. I have shut it down yesterday in order to prevent it to turn precious electrical energy into noise and heat. I've subscribed to the package varmon, so I will probably notice when it could be helpful to restart it. Perhaps I will reuse the computer's parts one day, but I plan to store the controller along with the system disks in a safe place, so I can revive it when it's needed. Whenever you would like to have it running again, write me an email. One more question: why does package depend on such recent ncurses and libc versions? | Die folgenden Pakete haben verletzte Abhängigkeiten: | varmon: | Hängt ab: libc6 (= 2.7-1) | aber 2.3.6.ds1-13etch7 ist installiert. | Hängt ab: libncurses5 (= 5.6+20071006-3) | aber 5.5-5 ist installiert. The binary runs perfectly on Etch (at least with backports), as far as I could test it, but it will show up as broken in the package management. Regards, Christoph signature.asc Description: PGP signature
Bug#491505: [package varmon] varmon segfaults on Etch i386
At 1218812561 time_t, Christoph Franzen wrote: the machine is set up and connected to the internet. Most of the things you need should already be installed. The system resides on a mirrored RAID Volume on the DAC960 with LVM. There are two additional disks for testing purposes. It's perfect, I can connect. I'll work probably tomorrow from work. I'll keep you in touch anyway. Thanks again, Cheers, -- Julien Danjou .''`. Debian Developer : :' : http://julien.danjou.info `. `' http://people.debian.org/~acid `- 9A0D 5FD9 EB42 22F6 8974 C95C A462 B51E C2FE E5CD signature.asc Description: Digital signature
Bug#491505: [package varmon] varmon segfaults on Etch i386
Hello guys, I finally decided to take a quick look this morning and I found the culprit. It was caused by *dead code* probably left in place by from previous upstream. I removed it, and now it works as expected in any gcc -O? mode. I'm preparing a new upstream release and will push it back to Debian ASAP (expect everything tomorrow). Thanks for the box Christoph. It was very useful. You can probably shut it down now. I'd suggest to keep it around just in case of a new bug in the coming weeks if you can, we never know, but that'd be odd. I've tested this new version with electric fence so I'm pretty confident. Cheers, -- Julien Danjou .''`. Debian Developer : :' : http://julien.danjou.info `. `' http://people.debian.org/~acid `- 9A0D 5FD9 EB42 22F6 8974 C95C A462 B51E C2FE E5CD signature.asc Description: Digital signature
Bug#491505: [package varmon] varmon segfaults on Etch i386
At 1218959576 time_t, Julien Danjou wrote: I finally decided to take a quick look this morning and I found the culprit. It was caused by *dead code* probably left in place by from previous upstream. I removed it, and now it works as expected in any gcc -O? mode. If people are interested I've pushed code here: git://git.naquadah.org/varmon.git http://git.naquadah.org/?p=varmon.git;a=summary The fix for this bug is: http://git.naquadah.org/?p=varmon.git;a=commitdiff;h=ae1cb95cfd553625eeaa9fa1f687c7181475fe8d Cheers, -- Julien Danjou // ᐰ [EMAIL PROTECTED] http://julien.danjou.info // 9A0D 5FD9 EB42 22F6 8974 C95C A462 B51E C2FE E5CD signature.asc Description: Digital signature
Bug#491505: [package varmon] varmon segfaults on Etch i386
Am Wed, 13 Aug 2008 16:27:45 +0200 schrieb Julien Danjou [EMAIL PROTECTED]: Hello Julien, All build-dep of varmon + gdb + eletric fence might help. Anyway I'll need root access to control the array so you can let me apt-get myself. the machine is set up and connected to the internet. Most of the things you need should already be installed. The system resides on a mirrored RAID Volume on the DAC960 with LVM. There are two additional disks for testing purposes. The main language is set to french, but the keymap is de-latin1-nodeadkeys (QWERTZ layout), because such a thing is attached to the machine. Feel free to change this into AZERTY or whatever you are using. I've got no internet flatrate here, so contact me before transferring huge amounts of data (shouldn't be necessary, I think). To save bandwidth I use the local Debian mirror of my ISP and relay this through my backup server with apt-cacher. This is transparent, so you can add repositories if needed. I will send you the password via encrypted mail soon. If you tell me which directories to copy I can also include these in my automated backup scheme via rsync, so nothing will be lost if the machine fails (the disks are rather old). Regards, Christoph -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#491505: [package varmon] varmon segfaults on Etch i386
Julien, Am Mon, 11 Aug 2008 22:35:12 +0200 schrieb Julien Danjou [EMAIL PROTECTED]: At 1218482627 time_t, Philipp Kern wrote: there is a Release Critical bug filed against your package. Unfortunately this package (varmon) needs special hardware to test it. Could you please look if the problem the submitter faces is currently reproduceable in Lenny? (The upstream version is the same, though.) No, because I don't have the hardware neither. I have got a spare DAC960PL Dual Channel Fast Wide SCSI RAID Controller with a few suitable disks, so I can build a Pentium II test box from that, install Lenny onto it, and give you SSH access. The box can be up permanently with full internet access via ADSL with a permanent IP address. However, I need a few days/weeks to set that up in my spare time. Please tell me if you are interested and provide the following information: 1) Your desired user name on the machine 2) Software packages that MUST be installed (specific versions needed?) 3) Software packages that MUST NOT be installed 4) Any wishes regarding the hardware 5) Anything else which would help you If you are interested I will maintain this machine as long as it is needed and I am able to. Besides, I speak also french (I mention this because your name sounds french to me). The etch version already had this problem (segfault) and it was avoided recompiling with a different -O option to gcc, see #401236. So I've no clue on how to debug and fix, and lenny may be affected anyway. It is the Etch version that I am using, and it HAS got the problem here. So I think the workaround doesn't help in all cases. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#491505: [package varmon] varmon segfaults on Etch i386
Hi Christoph, At 1218635679 time_t, Christoph Franzen wrote: However, I need a few days/weeks to set that up in my spare time. Please tell me if you are interested and provide the following information: 1) Your desired user name on the machine jd 2) Software packages that MUST be installed (specific versions needed?) All build-dep of varmon + gdb + eletric fence might help. Anyway I'll need root access to control the array so you can let me apt-get myself. 3) Software packages that MUST NOT be installed FreeBSD. ;) 4) Any wishes regarding the hardware amd64 might be better, but i386 will be fine. If you are interested I will maintain this machine as long as it is needed and I am able to. Besides, I speak also french (I mention this because your name sounds french to me). Yes, I'm french. I'm interested in fixing that problem. It won't take weeks, I'll let you know as soon as I find and fix it and you'll be able to unplug the box. :) Cheers, -- Julien Danjou .''`. Debian Developer : :' : http://julien.danjou.info `. `' http://people.debian.org/~acid `- 9A0D 5FD9 EB42 22F6 8974 C95C A462 B51E C2FE E5CD signature.asc Description: Digital signature
Bug#491505: [package varmon] varmon segfaults on Etch i386
Julien, there is a Release Critical bug filed against your package. Unfortunately this package (varmon) needs special hardware to test it. Could you please look if the problem the submitter faces is currently reproduceable in Lenny? (The upstream version is the same, though.) On Sun, Jul 20, 2008 at 12:58:21AM +0200, Christoph Franzen wrote: Varmon segfaults almost immediately. ~# varmon Scanning for VA safety backplane. Please wait a few moments... DAC960: Ctrlr 0, PCI 00:0b:01, IRQ 5, Channels 1 DAC960: Model DAC960PTL1, Firmware 4.08-0-37 Scanning Controller[0], Channel[0], ID[0] Segmentation fault I don't want to pull such specialised packages from testing lightly... Kind regards, Philipp Kern -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#491505: [package varmon] varmon segfaults on Etch i386
At 1218482627 time_t, Philipp Kern wrote: there is a Release Critical bug filed against your package. Unfortunately this package (varmon) needs special hardware to test it. Could you please look if the problem the submitter faces is currently reproduceable in Lenny? (The upstream version is the same, though.) No, because I don't have the hardware neither. The etch version already had this problem (segfault) and it was avoided recompiling with a different -O option to gcc, see #401236. So I've no clue on how to debug and fix, and lenny may be affected anyway. Cheers, -- Julien Danjou .''`. Debian Developer : :' : http://julien.danjou.info `. `' http://people.debian.org/~acid `- 9A0D 5FD9 EB42 22F6 8974 C95C A462 B51E C2FE E5CD signature.asc Description: Digital signature
Bug#491505: [package varmon] varmon segfaults on Etch i386
Package: varmon Version: 1.2.0-4 Severity: grave --- Please enter the report below this line. --- Varmon segfaults almost immediately. ~# varmon Scanning for VA safety backplane. Please wait a few moments... DAC960: Ctrlr 0, PCI 00:0b:01, IRQ 5, Channels 1 DAC960: Model DAC960PTL1, Firmware 4.08-0-37 Scanning Controller[0], Channel[0], ID[0] Segmentation fault --- System information. --- Architecture: i386 Kernel: Linux 2.6.25-2-686 Debian Release: etch A few Backports are installed. --- Package information. --- Depends: libc6 (= 2.3.6-6), libncurses5 (= 5.4-5) Filename: pool/main/v/varmon/varmon_1.2.0-4_i386.deb Depends(Version) | Installed -+-= libc6 (= 2.3.6-6) | 2.3.6.ds1-13etch2 libncurses5 (= 5.4-5) | 5.5-5 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]