Re: 'make all' in /sys/i386/boot fails

Sheldon Hearn wrote: > > On Fri, 25 Jun 1999 18:46:05 MST, Doug wrote: > > > Where would I go if I wanted to rebuild my boot blocks to make the > > com console run at a different speed, or is this something I could set > > with one of the voluminous conf files in /boot, or?? > > Try src/sys/boo

Re: 'make all' in /sys/i386/boot fails

On Sat, 26 Jun 1999 01:05:31 MST, Doug wrote: > Well that certainly looks like the right answer. Are you > uncertain about it, or just being rhetorical? The expanded version was Try src/sys/boot, which is where I found what looked very much like the new boot loader. I'v

Re: NT vs Linux benchmark saga continues (fwd)

And to follow up on this topic: the german magazine c't has an article in their current issue, where they compare a real-world setup measuring web-server performance for NT an Linux both for an SMP and non-SMP setup. Linux was - depending on the setup and application - a bit to much faster ...

Re: boot -P AND autoboot possible?

Sheldon Hearn wrote: > > On Fri, 25 Jun 1999 13:13:20 MST, Doug wrote: > > > I can submit patches for the man page(s) if I can get a grip on what's > > happening where (and when). :-/ What I'd really like to see is a > > chronological listing, like: > > I think you're reading the wrong manpage

Re: Out of file descriptors ??

On Fri, 25 Jun 1999, Scot W. Hetzel wrote: > What is happening is /etc/defaults/rc.conf pulls in /etc/rc.conf, which then > pulls in /etc/rc.conf repeat until we run out of file descriptors. I am not fond of the new defaults/rc.conf system. Used to be when I wanted to modify something in r

Re: place of logfile for cron (PR 7682) - move it?

It seems that no one really has any objections to moving the log file. These would the locations to change(find | grep /var/log) src/usr.sbin/cron/cron/config.h src/usr.sbin/cron/doc/CHANGES (src/usr.sbin/cron/doc/CHANGES.FreeBSD a la xntpd?) src/etc/Makefile src/etc/newsyslog.conf src/etc/sysl

Re: Out of file descriptors ??

On 26-Jun-99 Ben Rosengart wrote: > rc.conf, and change it -- unless I don't understand the new setup > correctly. That makes more work for me, not less. What's the idea > here? When you upgrade and use mergemaster its a lot less work.. The new defaults for the upgrade take place automatic

Re: Out of file descriptors ??

On Sat, 26 Jun 1999 09:21:59 GMT, Ben Rosengart wrote: > I am not fond of the new defaults/rc.conf system. Your objection is far too late, I'm afraid. The change makes staying STABLE and staying CURRENT much easier than it was before and has met with resounding approval, although it took a few

Re: ktrace causes kernel panic

Greg Lehey wrote: [snip] > >>> Using a recent (few days) -current I had a process lock up on me > >>> last night, so I did a 'ktrace -p whateverthepidwas' and let it run for a > >>> while. When I issued a 'ktrace -C' in another screen, everything froze and > >>> the kernel panic'ed. I dropped

-current breakage? (11.00 bst, 26/6/1999)

I've just tried building the world three times, cvsup'd at around 11.00 bst, and again at 12.00 and 13.00 - they all break at the point below... Any suggestions? :) - I was running "make buildworld -j8 | tee /var/tmp/mw.out". Each time I carefully blow away /usr/obj before re-running... -Kp " /

kernel profiling

Just curious: would it be easily doable to profile the kernel by sampling the program counter, à la profil(2) but for kernel mode? Has anyone attempted to write this? -- Pierre Beyssac[EMAIL PROTECTED] [EMAIL PROTECTED] {Free,Net,Open}BSD, Linux : il y a moins bien, mais c'est plus cher

USB project status

Project status of the USB project: Currently under development: Rewrite of uhub driver to take full advantage of the newbus architecture. It compiles but has not been tested yet. Rewrite of umass driver to take full advantage of CAM and its 'actions', using callbacks for US

Re: kernel profiling

At Sat, 26 Jun 1999 16:11:43 +0200, Pierre Beyssac <[EMAIL PROTECTED]> wrote: > > Just curious: would it be easily doable to profile the kernel by > sampling the program counter, la profil(2) but for kernel mode? > Has anyone attempted to write this? You can enable kernel profiling by configure

Re: kernel profiling

On Sun, 27 Jun 1999, Hidetoshi Shimokawa wrote: > At Sat, 26 Jun 1999 16:11:43 +0200, > Pierre Beyssac <[EMAIL PROTECTED]> wrote: > > > > Just curious: would it be easily doable to profile the kernel by > > sampling the program counter, la profil(2) but for kernel mode? > > Has anyone attempted

Who leads networking?

Hey Folks, I'm new to contributing to FreeBSD but not to working with kernels etc. I'm thinking of adding a new framing type to PPP which is described in the latest Transactions on Networking. Who do I talk to about this kind of stuff? Thanks, George To Unsubscribe: send mail to [E

kernel panic w/regard to setting io permissions bitmap

Hi, I'm chasing down a panic that occurs when you call 'i386_set_ioperm()' in -current (19990625-SNAP). This call worked correctly in the 19990426-SNAP. I first noticed the problem with the 6/10 snap and it's still with us as of the 6/25 snap. The panic occurs at 'swtch_com+0x6e: ltr %si'. At

Re: place of logfile for cron (PR 7682) - move it?

Nick Hibma once wrote: > It seems that no one really has any objections to moving the log file. > > These would the locations to change(find | grep /var/log) > > src/usr.sbin/cron/cron/config.h > src/usr.sbin/cron/doc/CHANGES > (src/usr.sbin/cron/doc/CHANGES.FreeBSD a la xntpd?) > src/etc/Makef

Re: kernel profiling

Doug Rabson <[EMAIL PROTECTED]> writes: > Another excellent alternative for kernel profiling is iprobe. Can we get a pointer? /assar To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message

Re: kernel profiling

In message <[EMAIL PROTECTED]>, Do ug Rabson writes: >> You can enable kernel profiling by configure kernel with `-p' or '-pp' >> option. See config(8) and kgmon(8) for detail. >> >> I have a patch for FreeBSD/alpha to support kernel profiling. > >Another excellent alternative for kernel profili

Re: kernel profiling

On 26 Jun 1999, Assar Westerlund wrote: > Doug Rabson <[EMAIL PROTECTED]> writes: > > Another excellent alternative for kernel profiling is iprobe. > > Can we get a pointer? There is a nice tutorial at http://www.cs.duke.edu/ari/iprobe.html. This tool is alpha specific though. -- Doug Rabson

Re: kernel profiling

On Sat, 26 Jun 1999, Poul-Henning Kamp wrote: > In message <[EMAIL PROTECTED]>, Do > ug Rabson writes: > > >> You can enable kernel profiling by configure kernel with `-p' or '-pp' > >> option. See config(8) and kgmon(8) for detail. > >> > >> I have a patch for FreeBSD/alpha to support kernel p

Building -current with 3.1?

Hi Folks, I've been getting the src-sys stuff via CVS and now want to build it. Can this be done with just the 3.1 tools (right off the CD) or do I need the compilers etc. as well? I get the following when doing a make in src/sys: ===> boot/i386/boot2 cc -elf -I/usr/local/FreeBSD-curr

Re: Building -current with 3.1?

Hello, The standard compiler has changed between 3.x and -Current (gcc2.7.2.1 => egcs) - so don't even think of compiling code from -Current with Stable tools. You may want to try some "make upgrade" to upgrade all of your OS from 3.X to 4.0 (then you will use the -Current sources for your kerne

Re: Building -current with 3.1?

> The standard compiler has changed between 3.x and -Current (gcc2.7.2.1 > => egcs) - so don't even think of compiling code from -Current with > Stable tools. > > You may want to try some "make upgrade" to upgrade all of your OS from > 3.X to 4.0 (then you will use the -Current sources for your k

Re: Building -current with 3.1?

On Sat, Jun 26, 1999 at 12:53:31PM -0700, George V. Neville-Neil wrote: > > The standard compiler has changed between 3.x and -Current (gcc2.7.2.1 > > => egcs) - so don't even think of compiling code from -Current with > > Stable tools. > > > > You may want to try some "make upgrade" to upgrade a

Ports problems

Hi, I just wondered if it was a problem with my installation or in general, but is anyone else here having problems with ports being dependant on glib12.2 and ports installing glib12.3 instead? I notice I cant install a LARGE number of the ports because of this problem, any suggestions? Andrew

Re: boot -P AND autoboot possible?

> > > I can submit patches for the man page(s) if I can get a grip on what's > > > happening where (and when). :-/ What I'd really like to see is a > > > chronological listing, like: > > > > I think you're reading the wrong manpage. I've just had a look at > > loader(8) and can't really see what

Re: Who leads networking?

In article <[EMAIL PROTECTED]>, George V. Neville-Neil <[EMAIL PROTECTED]> wrote: > > I'm new to contributing to FreeBSD but not to working with kernels > etc. I'm thinking of adding a new framing type to PPP which is described in > the latest Transactions on Networking. Who do I talk to

Re: 'make all' in /sys/i386/boot fails

> > Where would I go if I wanted to rebuild my boot blocks to make the > > com console run at a different speed, or is this something I could set > > with one of the voluminous conf files in /boot, or?? > > Try src/sys/boot ? That's the right place to go for the rebuild, but the setting is in /

Re: Who leads networking?

< said: > If you want the person specifically in charge of something, see > . For networking it's > Garrett Wollman <[EMAIL PROTECTED]>. And in the particular case of PPP, I defer to its maintainer, <[EMAIL PROTECTED]>. -GAWollman -- Garrett A.

COMMENTs ?

just cvsupping, why are all the ports getting a COMMENT file ? the DESCR file is not enough anymore ? == Tomer Weller [EMAIL PROTECTED] [EMAIL PROTECTED] "Drugs are good, and if you do'em pepole think that you're cool", NoFX To Unsubscribe: send mai

Re: COMMENTs ?

On Sun, 27 Jun 1999, Tomer Weller wrote: > just cvsupping, why are all the ports getting a COMMENT file ? the DESCR > file is not enough anymore ? > COMMENT and DESCR have always been there; see: http://www.freebsd.org/handbook/porting.html - Chris D. Faulhaber <[EMAIL PROTECTED]> | All th

Re: Out of file descriptors ??

> On Fri, 25 Jun 1999, Scot W. Hetzel wrote: > > > What is happening is /etc/defaults/rc.conf pulls in /etc/rc.conf, which then > > pulls in /etc/rc.conf repeat until we run out of file descriptors. > > I am not fond of the new defaults/rc.conf system. Used to be when I > wanted to modify

Re: 'make all' in /sys/i386/boot fails

> > > > sys/i386/boot is deprecated. > > Ok, well actually that could explain why it didn't work then. :) > Where would I go if I wanted to rebuild my boot blocks to make the > com console run at a different speed, or is this something I could set > with one of the voluminous conf files in

Re: kernel panic w/regard to setting io permissions bitmap

>The panic occurs at 'swtch_com+0x6e: ltr %si'. At the time, %si is >0x20, which indexes to what appears to be a busy tss (type=01011b) >with a limit of 0x67. That instruction generates a general protection >fault with an error code of 0. The current process was Idle. >... >Perhaps someone coul

Re: gdb no longer works?

This has been happening to me, as well! gdb takes a sig 11 and dies on coredumps on my -current as well. Can't even get to a (gdb) prompt! Also, gcore is busted! Barfs on a line from /proc/${PID}/map It appears that the format of /map lines has changed, and the scanf is failing... - Origina

Re: COMMENTs ?

On Sat, Jun 26, 1999 at 06:34:11PM -0400, Chris D. Faulhaber wrote: > On Sun, 27 Jun 1999, Tomer Weller wrote: > > > just cvsupping, why are all the ports getting a COMMENT file ? the DESCR > > file is not enough anymore ? > > COMMENT and DESCR have always been there; see: > http://www.freebsd.o

General feeling on merging APIs...

In message <[EMAIL PROTECTED]> "Justin T. Gibbs" writes: : The hope is to provide a consistent interface across : all *BSDs which is why I've addressed this to all of the *BSD projects. More generally, I'd like to identify areas where the kernel APIs of the various projects have diverged and see

Re: Changing the semantics of splsoftclock()

Bruce Evans wrote: > >>Why have splr semantics? That is, it raises to splsoftclock if current > >>priority is lower, else doesn't fiddle with it. > > splsoftclock() has always had spllower() semantics, and its main users > (kern_clock.c and kern_time.c) depend on this. > > FreeBSD has a preceden

Re: kernel panic w/regard to setting io permissions bitmap

>>The panic occurs at 'swtch_com+0x6e: ltr %si'. At the time, %si is >>0x20, which indexes to what appears to be a busy tss (type=01011b) >>with a limit of 0x67. That instruction generates a general protection >>fault with an error code of 0. The current process was Idle. The busy tss seems to