Re: considering i386 as a tier 1 architecture
Goodness gracious, did no one see the date on the original post? What's the limit on this fishing hole? -- Chris BeHanna ch...@behanna.org ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "freebsd-hackers-unsubscr...@freebsd.org"
Re: Where is FreeBSD going?
On Thursday 08 January 2004 13:05, Munish Chopra wrote: > On 2004-01-08 17:29 +, Doug Rabson wrote: > > [...] > > > The three main showstoppers for moving FreeBSD to subversion would be: > > > > 1. A replacement for cvsup. Probably quite doable using svnadmin > >dump and load. > > 2. Support for $FreeBSD$ - user-specified keywords are not supported > >and won't be until after svn-1.0 by the looks of things. > > 3. Converting the repository. This is a tricky one - I tried the > >current version of the migration scripts and they barfed and died > >pretty quickly. Still, I'm pretty sure that the svn developers > >are planning to fix most of those problems. From mailing-list > >archives, it appears that they are using our cvs tree as test > >material for the migration scripts. > > [...cvs2svn.py scheduled for 1.0...] What about "arch"? I have it installed, but $realjob has prevented me from looking at it. And, unless I misunderstand, Perforce is available for free for non-profits, and the client is a free download. Other than a desire to be "pure" and use open source exclusively, what objection is there to Perforce? (And even considering that desire, Perforce is built upon open source: RCS and BDB, if I understand correctly). Speaking as a former CVS repo-meister (for a company that evaporated out from under me), Perforce really is a better mousetrap. No more, "I updated in the middle of a commit" problem, because commits are transactional. No more "Oh, god, this merge sucks", because Perforce keeps track of what was merged when, and where. The latest versions rather painlessly support cross-branch merges, too (i.e., pulling changes from one branch to another without having to first push up to and pull down from a common ancestor). Triggers can be written to prevent inadvertent DoSes (p4 integ -I //depot/branch1/... //depot/branch2/...) and to do submit-time checks. Risks are more easily mitigated with branches, and pulling/pushing of selected changes is MUCH easier (no more need to generate and apply patches by hand). Generating weird-elmo hybrid mappings of the tree is also a snap, and the repo itself doesn't bloat as badly because P4 uses its database to keep track of where histories go, rather than actually physically copy a file to move it in the repo. CVS: cp /CVSROOT/foo/bar /CVSROOT/foo/baz cvs delete foo/bar cvs commit (but bar,v lives forever, if you want to keep the change history and/or if you ever want to check out an old tagged revision of the tree) Perforce: p4 integ -t foo/bar foo/baz p4 delete foo/bar p4 submit (foo/baz doesn't actually physically exist. P4 keeps a DB record that foo/baz points to foo/bar, and this operation is only visible in the branch in which it was done, until that branch is pushed up to its parent) With Perforce, no repo-meister intervention is needed. Add in the ability to use local proxies to cache frequently-fetched files and revisions, and you have a winner, IMHO. I'm starting to sound like a spokesman. I'm not--just a *very* satisfied user. -- Chris BeHanna Software Engineer (Remove "bogus" before responding.) [EMAIL PROTECTED] Turning coffee into software since 1990. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: running wine automatically as a shell for w32 binaries
On Saturday 06 December 2003 10:19, Julian Stacey wrote: > > Hi all, > > I wrote a new imgact function for FreeBSD to start wine > > automatically as a sort of an "interpreter" for windows > > binaries. > > <http://www.mawit.com/~jau/imgact_wine-4.9.patch> > > Great idea ! If this small diff gets tested & merged into src/ > automatic MS support will be a real plus. With the weekly proliferation of MS worms, trojans, and viruses, do you *really* think this is a good idea? Yeah, it's neat, nifty, and cool, but it comes with substantial risk. -- Chris BeHanna Software Engineer (Remove "bogus" before responding.) [EMAIL PROTECTED] Turning coffee into software since 1990. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Tuning HZ for semi-realtime applications
On Sunday 03 August 2003 16:54, Sean Hamilton wrote: > Greetings, > > [...wants to send out a lot of traffic, then read responses 1000 > times per second...is currently using select(2) in a loop...] > > Should I set HZ to 1000 (the frequency of my application) or should I set > it to a much higher value? The CPU is running at around 2 GHz, and I set it > as high as 50,000 with no problems. However, the granularity of my timeout > appears to be restricted to 1/1000th of a second. harti@ already answered this. I have no experience playing with HZ settings, but his response sounds reasonable enough. > I would like to use poll(2) instead of select, but it appears to take its > timeout parameter in milliseconds, which aren't precise enough to keep my > timing reasonable, especially if I ever need to increase my frequency. > > Another option would be calling poll/select with no timeout, in a loop. > However, this seems like a waste of CPU time. You could insert an appropriately-sized nanosleep(2) into such a loop. -- Chris BeHanna Software Engineer (Remove "bogus" before responding.) [EMAIL PROTECTED] Turning coffee into software since 1990. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: getfsent(3) and spaces in fstab
On Thu, 31 Jul 2003, Simon Barner wrote: > > Just a thort, not having tried it .. > > > > does either of the 'standard' methods of including spaces actually work > > in fstab ?? > > > > I speak of quoting (either single or double) and backslashing the space > > > > "/mnt/space/silly long dirname/filename also with spaces" > > > > or > > > > /mnt/space/silly\ long\ dirname/filename\ also\ with\ spaces > > Sorry, I should have written that I have performed tests: > > Here is what I did: > > test\ 1 /mnt/test\ 1ufs ro 0 0 > 'test 2''/mnt/test 2' ufs ro 0 0 > "test 3""/mnt/test 3" ufs ro 0 0 > > This test program > > [...snipped...] > > Gives me the following output: > > fstab: /etc/fstab:14: Inappropriate file type or format > fstab: /etc/fstab:15: Inappropriate file type or format > fstab: /etc/fstab:16: Inappropriate file type or format What about test%201/mnt/test%201 ufs ro 0 0 ? Ugly, yes, but that's how a lot of tools escape spaces. -- Chris BeHanna Software Engineer (Remove "bogus" before responding.) [EMAIL PROTECTED] Turning coffee into software since 1990. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: FreeBSD Auto-update (Was: Re: resolv and dynamic linking tocompatlibc)
On Tue, 2 Jul 2002, Colin Andrew Percival wrote: > [Apologies if this gets delivered twice; some broken DNS is causing mail > sent via shaw.ca to bounce.] > > At 10:32 02/07/2002 -0400, Chris BeHanna wrote: > >On Mon, 1 Jul 2002, Brett Glass wrote: > >> Alas, ethics demand that [older code which is now known to have security > >> flaws] be either taken offline or accompanied > >> with a clear, visible, and strong warning. > > > >Who is going to expend the time and effort to do this, and what > >task should they let drop on the floor to get it done? > > > >> A snapshot of 4.6-STABLE should also be made and released as 4.6.1. > > > > You could contribute to that, for a start, to make sure that the > >modularity needed to plug in an update facility is designed in. I'd > >suggest piggybacking the update facility on top of portupgrade to > >minimize duplication of effort. That, of course, depends upon the > >availability of known good binary packages with valid MD5 checksums > >and/or PGP signatures, and that's a whole 'nother resource problem. > > I'm new here (well, I've only been around for a bit over a year) so I'm > probably hopelessly lost, but... what is wrong with making world and > (GENERIC) kernel each time the 4.6 security branch is updated, and > publishing (signed) lists of the form "if you have file X with md5 hash > X_hash, replace it with file Y with md5 hash Y_hash" (where X is a local > path, and Y is a URL)? That's the basic idea, in part. If cons, for example, had the ability to use a URL to point to a cache, this could work just ducky (cons uses MD5 hashes rather than timestamps to determine if a file is out-of-date. That's a big win over make; however, cons can't yet do parallel builds). If the base system were itself divided into packages, then a solution based upon portupgrade could be employed. It requires someone to invest the time to set it up, and it requires some dedicated, trusted hardware, as you point out below: > I'd do this myself, except that I don't have any secure system to do this, > and I'd be horrified if anyone would trust binary updates coming from me > anyway. Another part of the puzzle is generating and supplying trusted precompiled packages from the ports tree. Finally, the last link in the chain is teaching sysinstall to automatically search for newer packages than were burned onto the CD, so that it can prompt the user to install the newer (presumably more secure) versions. Brett has been moaning for a very long time that this mechanism isn't in place, but he hasn't lifted a finger to help put it in place. -- Chris BeHanna http://www.pennasoft.com Principal Consultant PennaSoft Corporation [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
FreeBSD Auto-update (Was: Re: resolv and dynamic linking to compatlibc)
On Mon, 1 Jul 2002, Brett Glass wrote: > At 12:22 PM 7/1/2002, Jacques A. Vidrine wrote: > > >Gee, I guess we better get cracking to take offline every previous > >version of libc, too --- which would mean every version of FreeBSD and > >who knows what else. > > Alas, ethics demand that they be either taken offline or accompanied > with a clear, visible, and strong warning. Who is going to expend the time and effort to do this, and what task should they let drop on the floor to get it done? > [...snip...] > > A snapshot of 4.6-STABLE should also be made and released as 4.6.1. I repeat my above question. Yes, "Windows-update"-like machinery in sysinstall would be nice, but I don't see you falling all over yourself to either contribute it or fund it. All I see you doing is complaining that your priorities aren't adopted by the committers. What you want is more likely to occur if/when sysinstall is cleaned up. By all accounts, it's a maze of twisted little code paths, all alike. You could contribute to that, for a start, to make sure that the modularity needed to plug in an update facility is designed in. I'd suggest piggybacking the update facility on top of portupgrade to minimize duplication of effort. That, of course, depends upon the availability of known good binary packages with valid MD5 checksums and/or PGP signatures, and that's a whole 'nother resource problem. Followups to -hackers. -- Chris BeHanna http://www.pennasoft.com Principal Consultant PennaSoft Corporation [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: RE: FreeBSD vs Linux, Solaris, and NT
On Tue, 19 Dec 2000, Matt Dillon wrote: > Yes, it's a pretty sad state of affairs. What annoys me the most is > that companies actually believe they are protecting something when > they don't make their device driver source or hardware documentation > available. It has been well proven for years that the most withholding > accomplishes for the vast majority of these device drivers is a slight > delay--- perhaps a week or two, before competitors figure out what > they've done. Pirates don't care... they want the binaries anyway, > they aren't programmers. And the open-source community has always > strictly adhered to copyright and license restrictions. So all these > companies are doing is making life harder for themselves and for > their products. Unnecessarily. The XFree folks have some godaweful > stories about the crap they've had to wade through to get video > manufacturers on-board. Some video manufacturers have figured it out, > a lot haven't. Indeed. If anything, releasing specs and reference driver code should serve to increase the market for a given piece of hardware. The free OS market is into double digits now as a percentage of the total market. That's certainly not dominant, but it *is* significant. Personally, I will not buy any hardware that doesn't work on FreeBSD, period. That said, I find binary drivers acceptable, but I prefer open source drivers. I have some legacy hardware that isn't yet supported, that's waiting for a Round Tuit for me to write the needed code. When I've done that, I'm going to contribute it as my way of giving something back--a commercial OS as good as FreeBSD would be very, very costly, after all, and no one has charged me anything for it. > [...open source code is usually much better than closed commercial > code...] > > It's unfortunate, but there is light at the end of the tunnel. High > technology requires young minds and old managers are having a harder > and harder time dictating old paranoia to those people. If companies > want quality programmers they are having to become more flexible > and less paranoid. It is a slow process, but it is obviously working. The yound minds of today are the old minds of tomorrow. :-/ -- Chris BeHanna Software Engineer [EMAIL PROTECTED] Remove "bogus" before responding. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: ssh - are you nuts?!?
On Sat, 23 Dec 2000, David Preece wrote: > At 15:37 22/12/00 -0800, you wrote: > > >The question asked is: why you believe ssh is beter than say > >telnet. Or what advantages SSH has in general. > > Sorry, don't have time to reply to this properly. > > The main evil of ssh is that server authentication is not enforced, > making mounting a man-in-the-middle attack basically trivial. Man-in-the-middle or not, the fact that your data aren't transmitted in the clear automatically gives ssh a leg up over telnet, rsh, rlogin, and ftp. (At least one large company I know of has stated flatly, for example, that sending a root password over the wire in the clear is grounds for immediate termination.) You can certainly do your own server authentication, by carrying your known hosts file around on a floppy. ssh *does* warn you when you connect to a host that isn't present in your known hosts file--this isn't happening without your knowledge *and* consent. ssh may have its weaknesses, but telnet has little use other than as a diagnostic tool, IMHO (I only use it to send protocol commands to popd or sendmail these days). I'd *hardly* characterize ssh as "evil". -- Chris BeHanna Software Engineer [EMAIL PROTECTED] Remove "bogus" before responding. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: umask(2) and -Wconversion
On Tue, 7 Nov 2000, Peter Pentchev wrote:
> In my experience, the problem is not only with umask(2) - GCC *is*
> a bit stubborn about -Wconversion; I wonder if this is really a GCC bug :(
>
> I'm having the same problems with many other functions when passing
> integer constants - even if I explicitly cast them to a long or unsigned
> long or plain unsigned int or whatever the particular function needs,
> GCC seems to ignore the cast and whines about the conversion nonetheless :(
>
> Can anybody else confirm this? I can't dig out a code snippet right now,
> but ISTR a recurring case of this when compiling with BDECFLAGS a program
> which includes ncurses.h, then passes integer constants to init_pair()
> or something similar.
What happens if you pass an explicit long or unsigned long
literal instead of casting an integer literal? Like this:
void myfunc( unsigned long);
int
main( int argc, char* argv[])
{
myfunc( 12UL);
return 0;
}
I realize, of course, that going through and changing every array
subscript, for example, is a PITA. I'm just curious about whether or
not this makes gcc behave nicely.
--
Chris BeHanna
Software Engineer (at yourfit.com)
[EMAIL PROTECTED]
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message
Re: question for the freebsd community
On Wed, 25 Oct 2000, Michelle R. Sanchez, CNE wrote: > [...company makes high availability clustering software, and > supports FreeBSD...] > > we have had a lot of requests from customers wishing to make their > firewalls highly available by clustering them together and putting a > service monitor on the firewall port in case the firewall daemon should > hang. this is probably not very likely but they would like to be able to > do so in any case. > > my questions are these: > > 1] is it a good idea to try to put a service monitor on IPFW? If so, > does this compromise the firewall in any way? ipfw is not a daemon, and does not have a designated port to monitor--it's a kernel option to do packet filtering. If a kernel is built with the IPFIREWALL option, and the machine is running, then the firewall is also running, period. That should make the monitor as simple as asking the machine "Are you alive?". :-) I'd suggest "man ipfw" and also look at /sys/i386/conf/LINT for more details. -- Chris BeHanna Software Engineer (at yourfit.com) [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: gateway on different subnet
On Mon, 23 Oct 2000, Marko Ruban wrote:
> Summary of the problem:
>
> Can't assign cable modem gateway (10.17.56.12) to interface
> ed0 with assigned IP (208.59.162.242) - "network unreachable".
>
> I called RCN (my cable provider) and asked them to give me
> a gateway on the same subnet; they said they "don't do that".
man dhcp.conf(5). From what I saw there, "option routers" and
"option routes" are what you need to stick into your dhclient.conf.
Can you add routes by hand and make it work? If not, then even
the above won't help you.
> Part of solution:
>
> I set an alias for ed0 to 10.17.0.1 and it accepted the cable
> modem gateway as is. BUT, the packets are sent out with source
> address (10.17.0.1) responses to which, I suspect, gateway doesn't
> know how to route.
I would imagine you're right.
> Question: how can I have an alias of 10.17.0.1 and send out all
> packets with source address set to 208.59.162.242 (the IP
> that is actually assigned to the interface - not alias).
natd?
> P.S. Alternately, how can I force the system to allow a gateway
> that is on a different subnet (like windows allows that). Who
> can I turn to for help ?
I'm not a routing guru, but the guy in the next office is. I'll
ask him. (tromp, tromp, tromp)
The answer is this: run routed in verbose or debug mode, or
listen with rip, or some other network sniffer (you could even ping
the network and see who answers). You should have a local router
within 30 seconds.
Failing that, there have been a lot of other good suggestions,
among which probably the best is to switch to DSL. :-/
--
Chris BeHanna
Software Engineer (at yourfit.com)
[EMAIL PROTECTED]
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message
Re: can't build custom kernel
On Sat, 21 Oct 2000, Len Conrad wrote: > Hackers, > > Sorry to bother, but two postings to -question of the msg below in > two days got now response. I'm dead in the water. First thing: read /usr/src/UPDATING. The proper procedure to build a kernel is in there. To save you some time: cd /usr/src make buildkernel KERNEL= make installkernel KERNEL= If the build still fails, then yes, you have a legitimate problem. -- Chris BeHanna Software Engineer (at yourfit.com) [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: netgraph help(?)
On Sun, 8 Oct 2000, Brooks Davis wrote: > On Mon, Oct 09, 2000 at 11:59:44AM +1100, Tony Landells wrote: > > I'm trying to set up a multi-ethernet connection between two FreeBSD > > boxes so I can evaluate the performance of some of the network things > > (like IPSec) without bandwidth being an issue: > > No matter what you do that's going to add overhead. You'd certaintly > be better off with a pair of Gigabit ethernet adaptors. The netgear > ones are under $300US. Put 'em in shielded conduit. Do *NOT* grab a copper wire that is carrying oscillations at those frequencies! -- Chris BeHanna Software Engineer (at yourfit.com) [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: An sblock magic number is...
On Thu, 10 Aug 2000, Ollivier Robert wrote: > According to Wilko Bulte: > > Guessing: superblock? > > Yes. > > Our dump is only able to backup complete filesystems. Historically, SunOS' one > is able to backup partial FS (i.e. directories) and ironically, when Rémy Card > (of ext2fs fame) ported it to Linux, he added back that functionality. Short of that, this should work, shouldn't it? ssh user@host tar cvf - | gzip -c > myarchive.tgz If not, then pipe it to dd first, then gzip it. -- Chris BeHanna Software Engineer (at yourfit.com) [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: malloc to arrays?
On Fri, 4 Aug 2000, Aleksandr A.Babaylov wrote:
> Chris Costello writes:
> > On Wednesday, August 02, 2000, Mike Smith wrote:
> > > tcpcash_addr = (typeof(tcpcash_addr))
> > > malloc(sizeof(*tcpcash_addr) *
> > > TCPCASH_ROWSIZE *
> > > TCPCASH_COOLSIZE);
> >Just as a note on coding style, this will only hide warnings
> > caused by not including . malloc(3) returns void * and
> > does not need to be cast.
> So there is an error in gcc in base system?
> See:
>
> 1cicuta~/w/ra-tools(6)>cat tmp.c
> #include
>
> int main()
> { int *(i[100]);
>
> i = malloc(400);
> }
> 0cicuta~/w/ra-tools(7)>cc tmp.c
> tmp.c: In function `main':
> tmp.c:6: incompatible types in assignment
> 1cicuta~/w/ra-tools(8)>uname -a
> FreeBSD cicuta.babolo.ru 4.0-STABLE FreeBSD 4.0-STABLE #0: Fri Jun 9 14:01:52 MSD
>2000 [EMAIL PROTECTED]:/tmp/obj/usr/src/sys/cicuta i386
> 0cicuta~/w/ra-tools(9)>
There's no error at all. You've declared i to be an array of 100
pointers to int, but then you've tried to assign a pointer to that
array, and gcc quite naturally complains.
--
Chris BeHanna
Software Engineer (at yourfit.com)
[EMAIL PROTECTED]
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message
