Re: spin_lock on SCSI
Hi, backtrace available at http://www.net-amp.com/test/dump/ backtrace above is from fresh install with 2.8.2 CD SMP happens when starting heavy IO - in this case git pull :-( unfortunatly I did not compile my upgraded system with debugging enabled Matthew Dillon wrote: :Hi all, :any idea what I could do about this? : :HP WS Dual Xeon 1.7GHz, : :running UFS on da0, kernel SMP no APIC_IO, upgrade from dfly2.7xx :fs checked clean You have a crash dump. Is there a backtrace available to go along with that panic? -Matt
Re: MC not starting
I give up after fighting with pkgsrc for 2 hours. I tried compiling devel/libslang2 with pkgsrc's version of ncurses instead of native one. First I installed ncruses from pkgsrc, then changed mk.conf options to: LDFLAGS+= -Wl,-rpath -Wl,/usr/pkg/lib PREFER_PKGSRC= termcap ncurses Then I changed environment variables: # setenv LD_LIBRARY_PATH /usr/pkg/lib # setenv LD_RUN_PATH /usr/pkg/lib env shows that they're set properly. But after make replace in devel/libslang2, Slang is still using /usr/lib/libncurses.so.6. Something wrong on my site or libtool is broken on DragonFly BSD - ignoring LD_LIBRARY_PATH and so on? -- Paul Onyschuk bl...@bojary.koba.pl
Re: spin_lock on SCSI
:Hi, :backtrace available at :http://www.net-amp.com/test/dump/ :backtrace above is from fresh install with 2.8.2 CD SMP :... :happens when starting heavy IO - in this case git pull :-( Excellent, that backtrace was very useful. Here is an untested kernel source patch for you to try. You do need to compile up a kernel to test this. A spinlock was being held through a softupdates callback. The problem only occurs when running on UFS (and since most of us run on HAMMER we didn't notice). fetch http://apollo.backplane.com/DFlyMisc/vfs01.patch -Matt
Re: spin_lock on SCSI
thanx, will try it - give me some time to go through the build - coming back later Matthew Dillon wrote: :Hi, :backtrace available at :http://www.net-amp.com/test/dump/ :backtrace above is from fresh install with 2.8.2 CD SMP :... :happens when starting heavy IO - in this case git pull :-( Excellent, that backtrace was very useful. Here is an untested kernel source patch for you to try. You do need to compile up a kernel to test this. A spinlock was being held through a softupdates callback. The problem only occurs when running on UFS (and since most of us run on HAMMER we didn't notice). fetch http://apollo.backplane.com/DFlyMisc/vfs01.patch -Matt
comments.
Done testing 2.8.2. i´m using full 80Gb drive with joy. after using 2.6.3 for a month, the upgrades can be noticed. the filesystem consumes a bit more than 8 percent perhaps closer to 10% on registers while others like reiserfs consumes roughly 6% and ext3 3-5%. FS responsiveness is good, it´s writing perfomance is quite good given the 64 bits structure. FS management of hdd reminds me to reiserfs4 in particular. i feel it extreme solid for a beta. very impressive hammer failover on defective drives. we got some minor reboots using Xorg drivers and other minor reports. like syslogd crash related kernel dumps. I´ll try to get a port for ext3/4 and some other ports i need, kernel builds, vkernels. then i´ll post more.
Re: MC not starting
Paul Onyschuk wrote: I give up after fighting with pkgsrc for 2 hours. pkgsrc does a pretty good job of replacing the environment with a 'clean' one - so I'm not sure if the evnironment stuff would take effect - you can e.g. 'bmake -d A ...' to get more information than you could ever want about things.. also theres a PKG_DEBUG_LEVEL or similar to see the details - see also www.netbsd.org/docs/pkgsrc/ for sure perhaps just for tests it might make sense to move your system curses out of the way just to see what happens.. although, of course, ideally the 'proper' build should work.. I'm not well versed enough in the 'prefer pkgsrc' stuff to know how to debug it - I'd think: /usr/pkgsrc/mk/curses.buildlink3.mk /usr/pkgsrc/mk/curses.builtin.mk and maybe: /usr/pkgsrc/mk/defaults/options.description /usr/pkgsrc/mk/bsd.options.mk /usr/pkgsrc/mk/pkg-build-options.mk might be some things to check.. perhaps there's some bug there? another idea maybe is to try another curses-using package to see if it would build against your version - if so- perhaps the particular package makefile needs some kind of option set..
NVIDIA/nouveau
Hello! I'm new to the DFBSD. (I subscribed today). I use DragonFly for desktop and I'm satisfied with this system.There is only one thing upsets me - nvidia driver for my GeForce GTX 260. I tried this rep (git://gitweb.dragonflybsd.org/~corecode/nvidia.git) to make FreeBSD's drivers works under DFBSD and I tried to compile nouveau from pkgsrc/wip. Neither one nor the other failed. Please let me know if someone has successful experience of getting hardware acceleration in DFBSD?
ctrl alt backspace does not exit X server
Hi, I just finished installing X org from pkgsrc like the handbook describes, everything went well, except, when i press ctrl +alt+ backspace simultaneouslyto exit x windows, it does not exit, like it describes in the handbook to exit x. I don't know if i left something out or I have to add something to xorg.conf. Regards Edward
Re: ctrl alt backspace does not exit X server
X.org removed that feature few versions ago, if you want to have the ability to use it again need to start it with -retro. On Sat, Nov 6, 2010 at 01:05, Edward Martinez edwa...@bak.rr.com wrote: Hi, I just finished installing X org from pkgsrc like the handbook describes, everything went well, except, when i press ctrl +alt+ backspace simultaneously to exit x windows, it does not exit, like it describes in the handbook to exit x. I don't know if i left something out or I have to add something to xorg.conf. Regards Edward -- http://dfbsd.trackbsd.org.ar
Re: ctrl alt backspace does not exit X server
On Fri, Nov 5, 2010 at 5:37 PM, Sdävtaker sdavta...@gmail.com wrote: X.org removed that feature few versions ago, if you want to have the ability to use it again need to start it with -retro. Or, you can add the following to your xorg.conf: Section ServerFlags Option DontVTSwitch False Option DontZap False EndSection The first allows you to use CTRL+ALT+Function keys to switch between the console and X. The second allows you to use CTRL+ALT+Backspace to kill the X server. -- Freddie Cash fjwc...@gmail.com
Google Code-In! (details, instructions)
DragonFly was accepted as 1 of 20 organizations for Google Code-In! http://google-opensource.blogspot.com/2010/11/announcing-accepted-organizations-for.html We're 1 of 3 operating systems involved (us, Debian, and Haiku) and the only BSD involved. * The background: Code-In is like Summer of Code, in that it uses open source projects to assign work to volunteers in school, and Google pays the bill. Google Code-In is for 13 to 18-year-olds, as students. * If you aren't 13-18 years old: As described in the link above, sign up for an account at the Google Melange site: http://google-melange.com/ And request to be a mentor for DragonFly BSD. As a mentor, you can suggest projects for the students, and (if you want to) evaluate their work. We need project ideas that fit in the 8 different categories mentioned on the google-melange.com front page. You can never have too many ideas, so please contribute a few. Consider this an open invitation for everyone involved in DragonFly; not all the tasks have to be strictly code, so there isn't necessarily commit access involved. If you have friends involved in school teaching at the appropriate ages, let them know about this. * If you are 13-18 years old: Tasks become available for DragonFly and other projects November 22nd. Finishing 1 task gets you a T-shirt. Finishing 3 tasks gets you $100. You max out at $500 and 1 shirt. (15 tasks max compensation)