no serial console for ttyd0 HP Blade
Hi All, I installed a recent (as of today) RELENG_7 on one of our HP Blades. Unluckily my serial output stops right when a getty at ttyd0 should kick in. So I'm blind for the rest of the startup. Very unfortunate. The boot menu, as the bootup itself is printed out on serial console until: da0: Fixed Direct Access SCSI-5 device da0: 135.168MB/s transfers da0: 139947MB (286611840 512 byte sectors: 255H 32S/T 35124C) Trying to mount root from ufs:/dev/da0s1a bce0: link state changed to DOWN bce0: link state changed to UP bce1: link state changed to UP Which is, what I think, the point when init should take over and spawn a serial console on ttyd0, according to my /etc/ttys: # grep ttyd0 /etc/ttys ttyd0 "/usr/libexec/getty std.9600" vt100 on secure # cat /boot/loader.conf comconsole_speed="9600" console="vidconsole,comconsole" # A comma separated list of console(s) boot_multicons="-D" # -D: Use multiple consoles boot_serial="-h"# -h: Use serial console # uname -a FreeBSD db46-202.mobile.rz 7.0-STABLE FreeBSD 7.0-STABLE #0: Tue Jul 1 14:59:43 CEST 2008 [EMAIL PROTECTED]:/usr/obj/usr/src/sys/GENERIC amd64 The blade itself is a HP BL465c G5 Any ideas? I don't have this blade for a long time, so I'm a bit in a hurry. In fact, right now I'm testing this setup against a Debian 4.0 with Linux 2.6.25.9 under MySQL load, and up until now the machine looks good. If I wind up with a fully working, nearly as fast and stable as our linux boxes installation, I'd might be able to convince "the boys at work" to give FreeBSD a try. And in a 600 server setup, I'd be thrilled to do so :-) So lets get started with this serial console issue... which is indeed a pain :( TIA, Marian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: no serial console for ttyd0 HP Blade
This rhymes with sio deciding that your TTY is something other than ttyd0. We had this same problem on Rackable -- even though the proper tty was setup 0x3f8 irq 4 it was being assigned to sio1. You can see this by 'grep sio /var/log/messages' The only fix for this is to edit /boot/device.hints and reassign the flags to the sio1 interface, like so: hint.sio.1.at="isa" hint.sio.1.port="0x3F8" hint.sio.1.flags="0x10" hint.sio.1.irq="4" hint.sio.0.at="isa" hint.sio.0.port="0x2F8" hint.sio.0.irq="3" On Jul 1, 2008, at 8:23 AM, Marian Hettwer wrote: I installed a recent (as of today) RELENG_7 on one of our HP Blades. Unluckily my serial output stops right when a getty at ttyd0 should kick in. So I'm blind for the rest of the startup. Very unfortunate. The boot menu, as the bootup itself is printed out on serial console until: da0: Fixed Direct Access SCSI-5 device da0: 135.168MB/s transfers da0: 139947MB (286611840 512 byte sectors: 255H 32S/T 35124C) Trying to mount root from ufs:/dev/da0s1a bce0: link state changed to DOWN bce0: link state changed to UP bce1: link state changed to UP Which is, what I think, the point when init should take over and spawn a serial console on ttyd0, according to my /etc/ttys: # grep ttyd0 /etc/ttys ttyd0 "/usr/libexec/getty std.9600" vt100 on secure # cat /boot/loader.conf comconsole_speed="9600" console="vidconsole,comconsole" # A comma separated list of console(s) boot_multicons="-D" # -D: Use multiple consoles boot_serial="-h" # -h: Use serial console # uname -a FreeBSD db46-202.mobile.rz 7.0-STABLE FreeBSD 7.0-STABLE #0: Tue Jul 1 14:59:43 CEST 2008 [EMAIL PROTECTED]:/usr/obj/usr/src/sys/ GENERIC amd64 The blade itself is a HP BL465c G5 Any ideas? I don't have this blade for a long time, so I'm a bit in a hurry. In fact, right now I'm testing this setup against a Debian 4.0 with Linux 2.6.25.9 under MySQL load, and up until now the machine looks good. If I wind up with a fully working, nearly as fast and stable as our linux boxes installation, I'd might be able to convince "the boys at work" to give FreeBSD a try. And in a 600 server setup, I'd be thrilled to do so :-) So lets get started with this serial console issue... which is indeed a pain :( TIA, Marian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED] " -- Jo Rhett Net Consonance : consonant endings by net philanthropy, open source and other randomness ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: no serial console for ttyd0 HP Blade
Also remember that the blades have 2 serial ports, one can be accessed via a dongle in the front of the blade and I believe that is what usual would be called COM1 by default. The virtual serial via ilo takes up the COM2 by default. This is at least the default on DL series servers and I haven't checked into using the virtual serial on the Blades we have. On Tue, Jul 01, 2008 at 09:42:16PM -0700, Jo Rhett wrote: > This rhymes with sio deciding that your TTY is something other than > ttyd0. We had this same problem on Rackable -- even though the proper > tty was setup 0x3f8 irq 4 it was being assigned to sio1. You can see > this by 'grep sio /var/log/messages' > > The only fix for this is to edit /boot/device.hints and reassign the > flags to the sio1 interface, like so: > > hint.sio.1.at="isa" > hint.sio.1.port="0x3F8" > hint.sio.1.flags="0x10" > hint.sio.1.irq="4" > hint.sio.0.at="isa" > hint.sio.0.port="0x2F8" > hint.sio.0.irq="3" > > On Jul 1, 2008, at 8:23 AM, Marian Hettwer wrote: > >I installed a recent (as of today) RELENG_7 on one of our HP Blades. > >Unluckily my serial output stops right when a getty at ttyd0 should > >kick > >in. So I'm blind for the rest of the startup. > >Very unfortunate. > >The boot menu, as the bootup itself is printed out on serial console > >until: > >da0: Fixed Direct Access SCSI-5 device > >da0: 135.168MB/s transfers > >da0: 139947MB (286611840 512 byte sectors: 255H 32S/T 35124C) > >Trying to mount root from ufs:/dev/da0s1a > >bce0: link state changed to DOWN > >bce0: link state changed to UP > >bce1: link state changed to UP > > > >Which is, what I think, the point when init should take over and > >spawn a > >serial console on ttyd0, according to my /etc/ttys: > ># grep ttyd0 /etc/ttys > >ttyd0"/usr/libexec/getty std.9600" vt100 on secure > > > ># cat /boot/loader.conf > >comconsole_speed="9600" > >console="vidconsole,comconsole" # A comma separated list of > >console(s) > >boot_multicons="-D" # -D: Use multiple consoles > >boot_serial="-h" # -h: Use serial console > > > ># uname -a > >FreeBSD db46-202.mobile.rz 7.0-STABLE FreeBSD 7.0-STABLE #0: Tue > >Jul 1 > >14:59:43 CEST 2008 [EMAIL PROTECTED]:/usr/obj/usr/src/sys/ > >GENERIC > >amd64 > > > >The blade itself is a HP BL465c G5 > > > >Any ideas? I don't have this blade for a long time, so I'm a bit in a > >hurry. > >In fact, right now I'm testing this setup against a Debian 4.0 with > >Linux > >2.6.25.9 under MySQL load, and up until now the machine looks good. > >If I wind up with a fully working, nearly as fast and stable as our > >linux > >boxes installation, I'd might be able to convince "the boys at work" > >to > >give FreeBSD a try. > >And in a 600 server setup, I'd be thrilled to do so :-) > >So lets get started with this serial console issue... which is > >indeed a > >pain :( > > > >TIA, > >Marian > > > >___ > >freebsd-stable@freebsd.org mailing list > >http://lists.freebsd.org/mailman/listinfo/freebsd-stable > >To unsubscribe, send any mail to "[EMAIL PROTECTED] > >" > > -- > Jo Rhett > Net Consonance : consonant endings by net philanthropy, open source > and other randomness > > > ___ > freebsd-stable@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-stable > To unsubscribe, send any mail to "[EMAIL PROTECTED]" > -- Regards, Ulf. - Ulf Zimmermann, 1525 Pacific Ave., Alameda, CA-94501, #: 510-865-0204 You can find my resume at: http://www.Alameda.net/~ulf/resume.html ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: no serial console for ttyd0 HP Blade
I take that back, on blades the virtual serial is on COM1. On Tue, Jul 01, 2008 at 11:36:51PM -0700, Ulf Zimmermann wrote: > Also remember that the blades have 2 serial ports, one can be accessed via > a dongle in the front of the blade and I believe that is what usual would be > called COM1 by default. The virtual serial via ilo takes up the COM2 by > default. This is at least the default on DL series servers and I haven't > checked into using the virtual serial on the Blades we have. > > On Tue, Jul 01, 2008 at 09:42:16PM -0700, Jo Rhett wrote: > > This rhymes with sio deciding that your TTY is something other than > > ttyd0. We had this same problem on Rackable -- even though the proper > > tty was setup 0x3f8 irq 4 it was being assigned to sio1. You can see > > this by 'grep sio /var/log/messages' > > > > The only fix for this is to edit /boot/device.hints and reassign the > > flags to the sio1 interface, like so: > > > > hint.sio.1.at="isa" > > hint.sio.1.port="0x3F8" > > hint.sio.1.flags="0x10" > > hint.sio.1.irq="4" > > hint.sio.0.at="isa" > > hint.sio.0.port="0x2F8" > > hint.sio.0.irq="3" > > > > On Jul 1, 2008, at 8:23 AM, Marian Hettwer wrote: > > >I installed a recent (as of today) RELENG_7 on one of our HP Blades. > > >Unluckily my serial output stops right when a getty at ttyd0 should > > >kick > > >in. So I'm blind for the rest of the startup. > > >Very unfortunate. > > >The boot menu, as the bootup itself is printed out on serial console > > >until: > > >da0: Fixed Direct Access SCSI-5 device > > >da0: 135.168MB/s transfers > > >da0: 139947MB (286611840 512 byte sectors: 255H 32S/T 35124C) > > >Trying to mount root from ufs:/dev/da0s1a > > >bce0: link state changed to DOWN > > >bce0: link state changed to UP > > >bce1: link state changed to UP > > > > > >Which is, what I think, the point when init should take over and > > >spawn a > > >serial console on ttyd0, according to my /etc/ttys: > > ># grep ttyd0 /etc/ttys > > >ttyd0 "/usr/libexec/getty std.9600" vt100 on secure > > > > > ># cat /boot/loader.conf > > >comconsole_speed="9600" > > >console="vidconsole,comconsole"# A comma separated list of > > >console(s) > > >boot_multicons="-D"# -D: Use multiple consoles > > >boot_serial="-h" # -h: Use serial console > > > > > ># uname -a > > >FreeBSD db46-202.mobile.rz 7.0-STABLE FreeBSD 7.0-STABLE #0: Tue > > >Jul 1 > > >14:59:43 CEST 2008 [EMAIL PROTECTED]:/usr/obj/usr/src/sys/ > > >GENERIC > > >amd64 > > > > > >The blade itself is a HP BL465c G5 > > > > > >Any ideas? I don't have this blade for a long time, so I'm a bit in a > > >hurry. > > >In fact, right now I'm testing this setup against a Debian 4.0 with > > >Linux > > >2.6.25.9 under MySQL load, and up until now the machine looks good. > > >If I wind up with a fully working, nearly as fast and stable as our > > >linux > > >boxes installation, I'd might be able to convince "the boys at work" > > >to > > >give FreeBSD a try. > > >And in a 600 server setup, I'd be thrilled to do so :-) > > >So lets get started with this serial console issue... which is > > >indeed a > > >pain :( > > > > > >TIA, > > >Marian > > > > > >___ > > >freebsd-stable@freebsd.org mailing list > > >http://lists.freebsd.org/mailman/listinfo/freebsd-stable > > >To unsubscribe, send any mail to "[EMAIL PROTECTED] > > >" > > > > -- > > Jo Rhett > > Net Consonance : consonant endings by net philanthropy, open source > > and other randomness > > > > > > ___ > > freebsd-stable@freebsd.org mailing list > > http://lists.freebsd.org/mailman/listinfo/freebsd-stable > > To unsubscribe, send any mail to "[EMAIL PROTECTED]" > > > > -- > Regards, Ulf. > > - > Ulf Zimmermann, 1525 Pacific Ave., Alameda, CA-94501, #: 510-865-0204 > You can find my resume at: http://www.Alameda.net/~ulf/resume.html -- Regards, Ulf. - Ulf Zimmermann, 1525 Pacific Ave., Alameda, CA-94501, #: 510-865-0204 You can find my resume at: http://www.Alameda.net/~ulf/resume.html ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: no serial console for ttyd0 HP Blade
I'll bet you that sio is deciding that com1 or not, it's sio1 (not sio0) which can be fixed with the changes I mention below. On Jul 1, 2008, at 11:38 PM, Ulf Zimmermann wrote: I take that back, on blades the virtual serial is on COM1. On Tue, Jul 01, 2008 at 11:36:51PM -0700, Ulf Zimmermann wrote: Also remember that the blades have 2 serial ports, one can be accessed via a dongle in the front of the blade and I believe that is what usual would be called COM1 by default. The virtual serial via ilo takes up the COM2 by default. This is at least the default on DL series servers and I haven't checked into using the virtual serial on the Blades we have. On Tue, Jul 01, 2008 at 09:42:16PM -0700, Jo Rhett wrote: This rhymes with sio deciding that your TTY is something other than ttyd0. We had this same problem on Rackable -- even though the proper tty was setup 0x3f8 irq 4 it was being assigned to sio1. You can see this by 'grep sio /var/log/messages' The only fix for this is to edit /boot/device.hints and reassign the flags to the sio1 interface, like so: hint.sio.1.at="isa" hint.sio.1.port="0x3F8" hint.sio.1.flags="0x10" hint.sio.1.irq="4" hint.sio.0.at="isa" hint.sio.0.port="0x2F8" hint.sio.0.irq="3" On Jul 1, 2008, at 8:23 AM, Marian Hettwer wrote: I installed a recent (as of today) RELENG_7 on one of our HP Blades. Unluckily my serial output stops right when a getty at ttyd0 should kick in. So I'm blind for the rest of the startup. Very unfortunate. The boot menu, as the bootup itself is printed out on serial console until: da0: Fixed Direct Access SCSI-5 device da0: 135.168MB/s transfers da0: 139947MB (286611840 512 byte sectors: 255H 32S/T 35124C) Trying to mount root from ufs:/dev/da0s1a bce0: link state changed to DOWN bce0: link state changed to UP bce1: link state changed to UP Which is, what I think, the point when init should take over and spawn a serial console on ttyd0, according to my /etc/ttys: # grep ttyd0 /etc/ttys ttyd0 "/usr/libexec/getty std.9600" vt100 on secure # cat /boot/loader.conf comconsole_speed="9600" console="vidconsole,comconsole" # A comma separated list of console(s) boot_multicons="-D" # -D: Use multiple consoles boot_serial="-h" # -h: Use serial console # uname -a FreeBSD db46-202.mobile.rz 7.0-STABLE FreeBSD 7.0-STABLE #0: Tue Jul 1 14:59:43 CEST 2008 [EMAIL PROTECTED]:/usr/obj/usr/src/ sys/ GENERIC amd64 The blade itself is a HP BL465c G5 Any ideas? I don't have this blade for a long time, so I'm a bit in a hurry. In fact, right now I'm testing this setup against a Debian 4.0 with Linux 2.6.25.9 under MySQL load, and up until now the machine looks good. If I wind up with a fully working, nearly as fast and stable as our linux boxes installation, I'd might be able to convince "the boys at work" to give FreeBSD a try. And in a 600 server setup, I'd be thrilled to do so :-) So lets get started with this serial console issue... which is indeed a pain :( TIA, Marian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED] " -- Jo Rhett Net Consonance : consonant endings by net philanthropy, open source and other randomness ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED] " -- Regards, Ulf. - Ulf Zimmermann, 1525 Pacific Ave., Alameda, CA-94501, #: 510-865-0204 You can find my resume at: http://www.Alameda.net/~ulf/resume.html -- Regards, Ulf. - Ulf Zimmermann, 1525 Pacific Ave., Alameda, CA-94501, #: 510-865-0204 You can find my resume at: http://www.Alameda.net/~ulf/resume.html -- Jo Rhett Net Consonance : consonant endings by net philanthropy, open source and other randomness ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: no serial console for ttyd0 HP Blade
On Wed, Jul 02, 2008 at 12:16:41AM -0700, Jo Rhett wrote: > I'll bet you that sio is deciding that com1 or not, it's sio1 (not sio0) > which can be fixed with the changes I mention below. > > On Jul 1, 2008, at 11:38 PM, Ulf Zimmermann wrote: >> I take that back, on blades the virtual serial is on COM1. I agree with Joe. Chances are it may be "COM1" in the BIOS, but it may be actually wired to the equivalent of COM2. Or, the blade manufacturer set the ACPI table to point 0x3f8/4 to COM2 and 0x2f8/3 to COM1. I've seen this on one Supermicro board (front panel COM port is COM2, rear is COM1; but no mention of what's what in the manual. You actually have to experiment to find out.) -- | Jeremy Chadwickjdc at parodius.com | | Parodius Networking http://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, USA | | Making life hard for others since 1977. PGP: 4BD6C0CB | ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: no serial console for ttyd0 HP Blade
Hi there, On Tue, 1 Jul 2008 21:42:16 -0700, Jo Rhett <[EMAIL PROTECTED]> wrote: > This rhymes with sio deciding that your TTY is something other than > ttyd0. We had this same problem on Rackable -- even though the proper > tty was setup 0x3f8 irq 4 it was being assigned to sio1. You can see > this by 'grep sio /var/log/messages' > Well no, it seems the irq assignment for sioX is okay: Jul 1 15:13:57 db46-202 kernel: sio0: configured irq 3 not in bitmap of probed irqs 0 Jul 1 15:13:57 db46-202 kernel: sio0: port may not be enabled Jul 1 15:13:57 db46-202 kernel: sio0: configured irq 3 not in bitmap of probed irqs 0 Jul 1 15:13:57 db46-202 kernel: sio0: port may not be enabled Jul 1 15:13:57 db46-202 kernel: sio0: port 0x2f8-0x2ff irq 3 flags 0x10 on acpi0 Jul 1 15:13:57 db46-202 kernel: sio0: type 16550A, console Jul 1 15:13:57 db46-202 kernel: sio0: [FILTER] On the other hand... port may not be enabled doesn't sound good, hm? And in the end it gets irq3 at port 0x2f8 > The only fix for this is to edit /boot/device.hints and reassign the > flags to the sio1 interface, like so: > > hint.sio.1.at="isa" > hint.sio.1.port="0x3F8" > hint.sio.1.flags="0x10" > hint.sio.1.irq="4" > hint.sio.0.at="isa" > hint.sio.0.port="0x2F8" > hint.sio.0.irq="3" > Looks reasonable to me. I'll give it a shot. Thank you :) best regards, Marian > On Jul 1, 2008, at 8:23 AM, Marian Hettwer wrote: >> I installed a recent (as of today) RELENG_7 on one of our HP Blades. >> Unluckily my serial output stops right when a getty at ttyd0 should >> kick >> in. So I'm blind for the rest of the startup. >> Very unfortunate. >> The boot menu, as the bootup itself is printed out on serial console >> until: >> da0: Fixed Direct Access SCSI-5 device >> da0: 135.168MB/s transfers >> da0: 139947MB (286611840 512 byte sectors: 255H 32S/T 35124C) >> Trying to mount root from ufs:/dev/da0s1a >> bce0: link state changed to DOWN >> bce0: link state changed to UP >> bce1: link state changed to UP >> >> Which is, what I think, the point when init should take over and >> spawn a >> serial console on ttyd0, according to my /etc/ttys: >> # grep ttyd0 /etc/ttys >> ttyd0"/usr/libexec/getty std.9600" vt100 on secure >> >> # cat /boot/loader.conf >> comconsole_speed="9600" >> console="vidconsole,comconsole" # A comma separated list of >> console(s) >> boot_multicons="-D" # -D: Use multiple consoles >> boot_serial="-h" # -h: Use serial console >> >> # uname -a >> FreeBSD db46-202.mobile.rz 7.0-STABLE FreeBSD 7.0-STABLE #0: Tue >> Jul 1 >> 14:59:43 CEST 2008 [EMAIL PROTECTED]:/usr/obj/usr/src/sys/ >> GENERIC >> amd64 >> >> The blade itself is a HP BL465c G5 >> >> Any ideas? I don't have this blade for a long time, so I'm a bit in a >> hurry. >> In fact, right now I'm testing this setup against a Debian 4.0 with >> Linux >> 2.6.25.9 under MySQL load, and up until now the machine looks good. >> If I wind up with a fully working, nearly as fast and stable as our >> linux >> boxes installation, I'd might be able to convince "the boys at work" >> to >> give FreeBSD a try. >> And in a 600 server setup, I'd be thrilled to do so :-) >> So lets get started with this serial console issue... which is >> indeed a >> pain :( >> >> TIA, >> Marian >> >> ___ >> freebsd-stable@freebsd.org mailing list >> http://lists.freebsd.org/mailman/listinfo/freebsd-stable >> To unsubscribe, send any mail to "[EMAIL PROTECTED] >> " > > -- > Jo Rhett > Net Consonance : consonant endings by net philanthropy, open source > and other randomness > > > ___ > freebsd-stable@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-stable > To unsubscribe, send any mail to "[EMAIL PROTECTED]" ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: no serial console for ttyd0 HP Blade
Hi there, On Tue, 1 Jul 2008 23:36:51 -0700, Ulf Zimmermann <[EMAIL PROTECTED]> wrote: > Also remember that the blades have 2 serial ports, one can be accessed via > a dongle in the front of the blade and I believe that is what usual would > be > called COM1 by default. The virtual serial via ilo takes up the COM2 by > default. This is at least the default on DL series servers and I haven't > checked into using the virtual serial on the Blades we have. > Well, I enabled ttyd1 too in /etc/ttys and gave init a kill -HUP: [EMAIL PROTECTED] <~>ps ax | grep ttyd [9:54:02 on 08-07-02] 7888 ?? S 0:00.01 /usr/libexec/getty std.9600 ttyd1 961 d0 Is+0:00.00 /usr/libexec/getty std.9600 ttyd0 But no login prompt, though... regards, Marian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: no serial console for ttyd0 HP Blade
Hi Jeremy, On Wed, 2 Jul 2008 00:27:41 -0700, Jeremy Chadwick <[EMAIL PROTECTED]> wrote: > On Wed, Jul 02, 2008 at 12:16:41AM -0700, Jo Rhett wrote: >> I'll bet you that sio is deciding that com1 or not, it's sio1 (not sio0) >> which can be fixed with the changes I mention below. >> >> On Jul 1, 2008, at 11:38 PM, Ulf Zimmermann wrote: >>> I take that back, on blades the virtual serial is on COM1. > > I agree with Joe. Chances are it may be "COM1" in the BIOS, but it may > be actually wired to the equivalent of COM2. Or, the blade manufacturer > set the ACPI table to point 0x3f8/4 to COM2 and 0x2f8/3 to COM1. > Ack. > I've seen this on one Supermicro board (front panel COM port is COM2, > rear is COM1; but no mention of what's what in the manual. You actually > have to experiment to find out.) > I'll do so. However, I do wonder why the serial console works in boot loader and kernel bootup stage? Any explanation? :) regards, Marian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: no serial console for ttyd0 HP Blade
On Wed, 2 Jul 2008 00:16:41 -0700, Jo Rhett <[EMAIL PROTECTED]> wrote: > I'll bet you that sio is deciding that com1 or not, it's sio1 (not > sio0) which can be fixed with the changes I mention below. > puuuh, well, with your suggested change, I do get a login prompt now (at ttyd1 it says), but I don't see any bootup messages anymore. It seems that the sio0 is irq 4 while booting and becomes irq3 later. That's odd. If I remember correctly, one was able to configure the mapping of serial ports in the BIOS (in regards to those HP blades). Perhaps that helps. I'll give it a shot and if I found a way to have boot messages and the login getty, I'll drop a [solved] mail to this list. Anyway, thanks all so far :-) regards, Marian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: no serial console for ttyd0 HP Blade
On Jul 2, 2008, at 1:04 AM, Marian Hettwer wrote: puuuh, well, with your suggested change, I do get a login prompt now (at ttyd1 it says), but I don't see any bootup messages anymore. It seems that the sio0 is irq 4 while booting and becomes irq3 later. That's odd. If I remember correctly, one was able to configure the mapping of serial ports in the BIOS (in regards to those HP blades). Perhaps that helps. I'll give it a shot and if I found a way to have boot messages and the login getty, I'll drop a [solved] mail to this list. So the boot loader uses 0x3f8 irq 4 no matter what it's mapped to. Second stage does something similar, but After you've loaded the kernel it does mappings to sio0 and sio1 and these may be different. This is why you have to screw with device.hints so that all three stages are using the same device. Jul 1 15:13:57 db46-202 kernel: sio0: port 0x2f8-0x2ff irq 3 flags 0x10 on acpi0 It's pretty clear from your grep command that sio0 is 0x2f8 irq 3, but the boot loader always uses 0x3f8 irq 4. You have to change these until the device which has 0x3f8 irq 4 has the "flags 0x10" option, which is what makes it the console. NOTE: I think this whole parade sucks, and the kernel should believe device.hints ... but there is no apparent interest in solving this problem (even when a bounty is offered), and I don't do enough device driver development these days for it to be worth my time. -- Jo Rhett Net Consonance : consonant endings by net philanthropy, open source and other randomness ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
