About building release for using sysinstall
Good morning,
I have launched :
cd /usr/src
make -j3 buildworld (I'm in testing env, I say because -j3)
cd /usr/src/release
make -f Makefile.sysinstall release CHROOTDIR=/expert/RELENG90RELEASE
CVSROOT=/expert/ncvs RELEASETAG=RELENG_9_0 MAKE_ISOS=1
does not work out of the box directly. It gives some errors in the iso
generation part, and some warning because the own Makefile.sysinstall
does not find install.cfg.
Using the mkisoimages.sh file from RELENG_9_0 have seen that (unless
using Makefile.sysinstall with make) make release crashes because
mkisoimages.sh script tries to create an fstab file inside an /etc dir
which should be inside {$CHROOTDIR}/R/cdrom/discX/ but which is not out
of the box (as a consequence of the make release I mean unless with as I
said Makefile.sysinstall)... so in a previous instance... I modified
this script (mkisoimages.sh) for generating this etc dir in required
location and before generating fstab by this script. Release finished
this way building but CD Disc1 was not able to boot because crashed
(can't see exactly the error, it's pretty fast when appears although I
know it's in mounting root phase, and I'm not using kernel debugging
config, and the machine reboots instantly) but was solved removing this
/etc dir from {$CHROOTDIR}/R/cdrom/discX/ and regenerating the iso file
again... and makes lots of sense... because in loader.conf you are
telling to boot and use as root an mfsroot not the own cd9660 image
which is what you're specifying in /etc/fstab in the iso image. So
finally this is what has worked to me for generating the iso images of
the new release :
I have no need of packages cd... so I removed it... and same for DVD
1... because I just need Disc1 and livefs (althought I have created too
the bootonly image...)... have seen too that doc distribution is copied
now to cd1 and dvd1...
Like have a local copy of FreeBSD cvs have made a couple of commits
which have allowed me building the release and ISOS to work properly for
later being able to use for our unattended installation system. I paste
out here the cvs diffs :
generarelease90# cvs diff -r 1.5.2.1.2.2 -r 1.5.2.1.2.5
Makefile.sysinstall
Index: Makefile.sysinstall
===
RCS file: /expert/ncvs/src/release/Attic/Makefile.sysinstall,v
retrieving revision 1.5.2.1.2.2
retrieving revision 1.5.2.1.2.5
diff -r1.5.2.1.2.2 -r1.5.2.1.2.5
1c1
< # $FreeBSD: src/release/Makefile.sysinstall,v 1.5.2.1.2.2 2012/01/02
04:11:22 kensmith Exp $
---
# $FreeBSD: src/release/Makefile.sysinstall,v 1.5.2.1.2.5 2012/02/26
11:51:25 portero Exp $
525c525
<-test -f install.cfg && cp install.cfg ${CHROOTDIR}/usr/src/release
---
-test -f /usr/src/usr.sbin/sysinstall/install.cfg && cp
/usr/src/usr.sbin/sysinstall/install.cfg ${CHROOTDIR}/usr/src/release
838,839c838,839
<-test -f ${.CURDIR}/install.cfg \
<&& cp ${.CURDIR}/install.cfg ${RD}/mfsfd
---
-test -f /usr/src/usr.sbin/sysinstall/install.cfg \
&& /usr/src/usr.sbin/sysinstall/install.cfg ${RD}/mfsfd
1152,1155d1151
<@sh ${.CURDIR}/${TARGET}/mkisoimages.sh \
$1/etc/fstab
44d42
< rm $1/etc/fstab
If you see no problems on this changes... perhaps would be nice to
apply something like this to RELENG_9_0 in some way it does not break
make release with bsdinstall (so default Makefile) but in the way it
works for generating release and iso images for being able to use with
sysinstall (unless for having some time to change machine provisioning
systems)
What do you think about all this?.
Thanks a lot for you're time,
Best regards.
___
freebsd-hackers
Re: [RFT][patch] Scheduling for HTT and not only
2012/2/27 George Mitchell : > I finally got around to trying this on a 9.0-STABLE GENERIC kernel, in > the forlorn hope that it would fix SCHED_ULE's poor performance for > interactive processes with a full load on interactive processes. It > doesn't help. -- George Mitchell Are you using sysmouse (moused) for the xorg pointer ? -- Olivier Smedts _ ASCII ribbon campaign ( ) e-mail: oliv...@gid0.org - against HTML email & vCards X www: http://www.gid0.org - against proprietary attachments / \ "Il y a seulement 10 sortes de gens dans le monde : ceux qui comprennent le binaire, et ceux qui ne le comprennent pas." ___ 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: [RFT][patch] Scheduling for HTT and not only
On 02/27/12 05:35, Olivier Smedts wrote: 2012/2/27 George Mitchell: I finally got around to trying this on a 9.0-STABLE GENERIC kernel, in the forlorn hope that it would fix SCHED_ULE's poor performance for interactive processes with a full load on interactive processes. It doesn't help. -- George Mitchell Are you using sysmouse (moused) for the xorg pointer ? Yes.-- George Mitchell ___ 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: [RFT][patch] Scheduling for HTT and not only
2012/2/27 George Mitchell : > On 02/27/12 05:35, Olivier Smedts wrote: >> >> 2012/2/27 George Mitchell: >>> >>> I finally got around to trying this on a 9.0-STABLE GENERIC kernel, in >>> the forlorn hope that it would fix SCHED_ULE's poor performance for >>> interactive processes with a full load on interactive processes. It >>> doesn't help. -- George Mitchell >> >> >> Are you using sysmouse (moused) for the xorg pointer ? >> > > Yes. -- George Mitchell Can you try with hald, or directly with the mouse device, without using moused ? Others reported they had better interactivity without sysmouse/moused. Really better (no mouse lag or freeze when under high load). -- Olivier Smedts _ ASCII ribbon campaign ( ) e-mail: oliv...@gid0.org - against HTML email & vCards X www: http://www.gid0.org - against proprietary attachments / \ "Il y a seulement 10 sortes de gens dans le monde : ceux qui comprennent le binaire, et ceux qui ne le comprennent pas." ___ 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: SuperMicro IPMI keyboard - fails for 'mountroot>' prompt under FreeBSD 9-R...
--On 27 February 2012 12:04 + Karl Pielorz wrote: Once the kernel is loading you see: Sorry - that should be, once the O/S is booting, not kernel loading - you see: ... " ugen1.2: at usbus1 ums0: on usbus1 ums0: 3 buttons and [Z] coordinates ID=0 ukbd0: on usbus1 kbd2 at ukbd0 " -Karl ___ 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"
SuperMicro IPMI keyboard - fails for 'mountroot>' prompt under FreeBSD 9-R...
Hi, We have a number of SuperMicro based systems (e.g. an X8DTL-IF) - this is running the latest BIOS (2.1a) - and the latest (supplied by SuperMicro) IPMI firmware - 2.44. We're using FreeBSD 9.0-R amd64. When using the LAN KVM keyboard (IPMI) - it works for the BIOS (as you'd expect) - it works for the Beastie boot menu prompt, and it works for when the O/S is booted. However, it doesn't work if the boot fails, e.g. if you get stuck at: mountroot> I've tried changing a few things in the BIOS (such as EHCI handoff, legacy port 60/64 support, legacy USB support) - but nothing I change in there makes any difference. Once the kernel is loading you see: " ugen1.2: at usbus1 ums0: 1.10/0.01, addr 2> on usbus1 ums0: 3 buttons and [Z] coordinates ID=0 ukbd0: 1.10/0.01, addr 2?> on usbus1 kbd2 at ukbd0 " I've tried with a PS2 keyboard attached, and removed from the physical server - but it doesn't change anything. We have notes here for FreeBSD 8 to recompile the kernel with: nodevice ums To fix a previous KVM issue with SuperMicro - but this doesn't fix the above issue. Any clues on how I can fix / troubleshoot this further? The keyboard does work at the Beastie boot menu, and again once the O/S is loaded - but obviously having it work when the kernel has failed to load, would be handy :) Thanks, -Karl ___ 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: [RFT][patch] Scheduling for HTT and not only
on 27/02/2012 13:28 Olivier Smedts said the following: > Can you try with hald, or directly with the mouse device, without > using moused ? Others reported they had better interactivity without > sysmouse/moused. Really better (no mouse lag or freeze when under high > load). > I wonder if re-nice-ing moused to a higher priority would help as well... -- Andriy Gapon ___ 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: BUG: 9.0 stage 2 boot (/boot/boot)
On Saturday, February 25, 2012 9:41:48 am rank1see...@gmail.com wrote: > > Do you only see the "No " message? Do you see the '/boot.config: /loader' > > message? (Do you have RBX_QUIET enabled perhaps? (-q)) Do you get the > > actual > > boot2 prompt at all? > > I don't have RBX_QUIET enabled nor any other flags > > Let the pic tell a story: > http://www.starforce.biz/stage2boot.jpg Ahh, this is helpful. You do see the '/boot.config: /loader' message. > It is also valid for your latest patch > > > > Hmm, I think the problem is that 'opts' has garbage instead of being > > initialized to zero. > > > > Try this (also at www.freebsd.org/~jhb/patches/boot2_opts.patch): > > > Patch eliminates possible error, of manual "intervention" > That is, a perfectly valid patch being classified as invalid. I have no idea what you mean here. However, it seems you don't have junk in your 'opts' variable anyway. Hmm, you could try adding some more debugging to boot2.c to see exactly what is failing. For example, does the first call to 'parse()' fail and clear autoboot? Oh, garbage in cmd[] could be bad. We might read beyond the end of the file. Try http://www.FreeBSD.org/~jhb/patches/boot2_cmd.patch -- John Baldwin ___ 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: Kernel stalling at "pci0: on pcib0"
On Friday, February 24, 2012 5:42:08 pm Will McCutcheon wrote: > Hello everybody, > > I originally posted this at freebsd-questions and was referred over here. > > I recently got a HP t5700 thin client that I wanted to turn into a > firewall using pfSense. For reference, this system uses a Transmeta Crusoe > TM5800 CPU with a VIA chipset that I'm having difficulty identifying. My > issue is that about half of the time the kernel will stall during startup > after the following: > > Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 > The Regents of the University of California. All rights reserved. > FreeBSD is a registered trademark of The FreeBSD Foundation. > FreeBSD 9.0-RELEASE #0: Tue Jan 3 07:15:25 UTC 2012 > root at obrian.cse.buffalo.edu:/usr/obj/usr/src/sys/CENERIC i386 > CPU: Transmeta(tm) Crusoe(tm) Processor TM5800 (997.69-MHz 586-class CPU) > Origin = "CenuineTMx86" Id = 0x543 Family = 5 Model = 4 Stepping = 3 > Features= 0x84893f > real memory = 270532608 (258 MB) > avail memory = 226930688 (216 MB) > kbd1 at kbdmux0 > acpi0: on motherboard > acpi0: Power Button (fixed) > acpi0: Sleep Button (fixed) > acpi0: reservation of 0, a (3) failed > acpi0: reservation of 10, eef (3) failed > acpi_timer0: couldn't allocate resource (port 0x4008) > cpu0: on acpi0 > acpi_button0: on acpi0 > pcib0: port > Oxcf8-0xcff,0x4000-0x407f,0x4080-0x40ff,0x500 > 0-0x500f on acpi0 > pcib0: Length mismatch for 4 range: f00 vs eff > pcib0: Length mismatch for 4 range: aff0 vs afef > pciO: on pcib0 > > If I power the system off and on again there's about a 50/50 chance it > will start up properly, whereupon it will run for days without issue. It's > just during startup that I see any problems. I've tried two separate > systems of the same model but they both exhibit the same issue, thus > suggesting it's not a one-off hardware defect. I tried both pfSense 2.0.1 > (which is FreeBSD 8.1-based) and FreeBSD 9.0-RELEASE from a USB stick, but > they exhibited the same issue. > > I do see some acpi0 and pcib0 messages in there that seem possibly > problematic, but I'm a bit out of my depth here. Could anyone spare a > moment to suggest any further troubleshooting steps I might try? Thanks so > much for your time! Hmm, the 'Length mismatch' messages are interesting. Can you do a 'boot -v'? Also, try setting 'debug.acpi.disable=hostres' at the loader prompt. -- John Baldwin ___ 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: improving VM - questions
on 25/02/2012 17:45 Wojciech Puchar said the following: [snip] > i tried that patch > > --- swap_pager.c.orig 2012-02-25 16:22:25.0 +0100 > +++ swap_pager.c2012-02-25 13:19:51.0 +0100 > @@ -119,7 +119,7 @@ > * The 32-page limit is due to the radix code (kern/subr_blist.c). > */ > #ifndef MAX_PAGEOUT_CLUSTER > -#define MAX_PAGEOUT_CLUSTER 16 > +#define MAX_PAGEOUT_CLUSTER 256 > #endif [snip] > but swap_pager.c patch seems not to work. i observe 64kB pageouts, no more. > > what is wrong in it? Could there be a problem because of what the MAX_PAGEOUT_CLUSTER comment says? -- Andriy Gapon ___ 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: [RFT][patch] Scheduling for HTT and not only
On 02/27/12 06:28, Olivier Smedts wrote: 2012/2/27 George Mitchell: On 02/27/12 05:35, Olivier Smedts wrote: 2012/2/27 George Mitchell: I finally got around to trying this on a 9.0-STABLE GENERIC kernel, in the forlorn hope that it would fix SCHED_ULE's poor performance for interactive processes with a full load on interactive processes. It doesn't help. -- George Mitchell Are you using sysmouse (moused) for the xorg pointer ? Yes.-- George Mitchell Can you try with hald, or directly with the mouse device, without using moused ? Others reported they had better interactivity without sysmouse/moused. Really better (no mouse lag or freeze when under high load). This seems to be harder than you would expect with a USB mouse -- moused starts up even with 'moused_enable="NO"' in my /etc/rc.conf. And in any case, I'm not talking about poor mouse response: I'm talking about over a minute and a half for thunderbird to start up with the mouse standing still, and "make buildkernel" not completing even after eight hours. -- George Mitchell ___ 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"
