I have this odd issue with the 2.6.9 & greater kernels, that I can't ssh
or use scp any more, this is what happens when I try:
SSH Error with-in X:
ssh_askpass: exec(/usr/libexec/ssh-askpass): No such file or directory
Host key verification failed.
SCP Error with-in X:
ssh_askpass: exec(/usr/libexec/ssh-askpass): No such file or directory
Host key verification failed.
lost connection
I just get Host key verification failed in the terminal with either
command.
ssh-askpass does not exsisit but it never has & ssh/scp works fine with
2.6.8.1 and below. When upgrading to the new kernel I just copied my
old .config and did a make oldconfig, make, etc.
Any ideas?
Thanks,
Dan.
On Wed, 27 Oct 2004, Danny Brow wrote:
> That was it, thanks.
>
Ok, thank you for the feedback, glad you fixed your problem.
Now I guess we just need for someone to find out why LEGACY_PTYS breaks
ssh (and other apps?) with kernels >= 2.6.9, but I'm afraid thats beyond
my capabilities.
--
Jesper Juhl
I'm recompiling right now.
On Wed, 2004-10-27 at 22:54 +0200, Jesper Juhl wrote:
> On Wed, 27 Oct 2004, Danny Brow wrote:
>
> > I have this odd issue with the 2.6.9 & greater kernels, that I can't ssh
> > or use scp any more, this is what happens when I try:
> >
> > SSH Error with-in X:
> > ssh_askpass: exec(/usr/libexec/ssh-askpass): No such file or directory
> > Host key verification failed.
> >
> > SCP Error with-in X:
> > ssh_askpass: exec(/usr/libexec/ssh-askpass): No such file or directory
> > Host key verification failed.
> > lost connection
> >
> > I just get Host key verification failed in the terminal with either
> > command.
> >
> > ssh-askpass does not exsisit but it never has & ssh/scp works fine with
> > 2.6.8.1 and below. When upgrading to the new kernel I just copied my
> > old .config and did a make oldconfig, make, etc.
> >
> > Any ideas?
> >
>
> As it happens, I had a chat with a few people on IRC a few days ago who
> had very similar trouble and I had no such trouble. After talking about
> our kernel configs a bit we came to the conclusion that the main
> difference between our .config's (except for some hardware specific
> drivers and a few other bits that seemed unrelated) was that they had
> CONFIG_LEGACY_PTYS=y while I had it unset. One guy recompiled his kernel
> on the spot and disabled CONFIG_LEGACY_PTYS and then reported that his
> trouble went away. He may have made other changes as well, I don't know,
> and it may have been some of the seemingly unrelated bits that did the
> trick, again I don't know.
>
> I have not attempted to verify this myself, but I'd say it's worth a try
> for you to disable CONFIG_LEGACY_PTYS (if you have it enabled) and then
> test that (and report your findings back).
>
> --
> Jesper Juhl
That was it, thanks.
On Wed, 2004-10-27 at 17:25 -0400, Danny Brow wrote:
> I'm recompiling right now.
>
> On Wed, 2004-10-27 at 22:54 +0200, Jesper Juhl wrote:
> > On Wed, 27 Oct 2004, Danny Brow wrote:
> >
> > > I have this odd issue with the 2.6.9 & greater kernels, that I can't ssh
> > > or use scp any more, this is what happens when I try:
> > >
> > > SSH Error with-in X:
> > > ssh_askpass: exec(/usr/libexec/ssh-askpass): No such file or directory
> > > Host key verification failed.
> > >
> > > SCP Error with-in X:
> > > ssh_askpass: exec(/usr/libexec/ssh-askpass): No such file or directory
> > > Host key verification failed.
> > > lost connection
> > >
> > > I just get Host key verification failed in the terminal with either
> > > command.
> > >
> > > ssh-askpass does not exsisit but it never has & ssh/scp works fine with
> > > 2.6.8.1 and below. When upgrading to the new kernel I just copied my
> > > old .config and did a make oldconfig, make, etc.
> > >
> > > Any ideas?
> > >
> >
> > As it happens, I had a chat with a few people on IRC a few days ago who
> > had very similar trouble and I had no such trouble. After talking about
> > our kernel configs a bit we came to the conclusion that the main
> > difference between our .config's (except for some hardware specific
> > drivers and a few other bits that seemed unrelated) was that they had
> > CONFIG_LEGACY_PTYS=y while I had it unset. One guy recompiled his kernel
> > on the spot and disabled CONFIG_LEGACY_PTYS and then reported that his
> > trouble went away. He may have made other changes as well, I don't know,
> > and it may have been some of the seemingly unrelated bits that did the
> > trick, again I don't know.
> >
> > I have not attempted to verify this myself, but I'd say it's worth a try
> > for you to disable CONFIG_LEGACY_PTYS (if you have it enabled) and then
> > test that (and report your findings back).
> >
> > --
> > Jesper Juhl
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
On Wed, 2004-10-27 at 17:25 -0400, Danny Brow wrote:
> I'm recompiling right now.
>
> On Wed, 2004-10-27 at 22:54 +0200, Jesper Juhl wrote:
> > On Wed, 27 Oct 2004, Danny Brow wrote:
> >
> > > I have this odd issue with the 2.6.9 & greater kernels, that I can't ssh
> > > or use scp any more, this is what happens when I try:
> > >
> > > SSH Error with-in X:
> > > ssh_askpass: exec(/usr/libexec/ssh-askpass): No such file or directory
> > > Host key verification failed.
> > >
> > > SCP Error with-in X:
> > > ssh_askpass: exec(/usr/libexec/ssh-askpass): No such file or directory
> > > Host key verification failed.
> > > lost connection
> > >
> > > I just get Host key verification failed in the terminal with either
> > > command.
> > >
> > > ssh-askpass does not exsisit but it never has & ssh/scp works fine with
> > > 2.6.8.1 and below. When upgrading to the new kernel I just copied my
> > > old .config and did a make oldconfig, make, etc.
> > >
> > > Any ideas?
> > >
> >
> > As it happens, I had a chat with a few people on IRC a few days ago who
> > had very similar trouble and I had no such trouble. After talking about
> > our kernel configs a bit we came to the conclusion that the main
> > difference between our .config's (except for some hardware specific
> > drivers and a few other bits that seemed unrelated) was that they had
> > CONFIG_LEGACY_PTYS=y while I had it unset. One guy recompiled his kernel
> > on the spot and disabled CONFIG_LEGACY_PTYS and then reported that his
> > trouble went away. He may have made other changes as well, I don't know,
> > and it may have been some of the seemingly unrelated bits that did the
> > trick, again I don't know.
> >
> > I have not attempted to verify this myself, but I'd say it's worth a try
> > for you to disable CONFIG_LEGACY_PTYS (if you have it enabled) and then
> > test that (and report your findings back).
> >
This was the problem.
Thanks,
Dan.
> > --
> > Jesper Juhl
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
Jesper Juhl <[email protected]> wrote:
>
> Now I guess we just need for someone to find out why LEGACY_PTYS breaks
> ssh (and other apps?) with kernels >= 2.6.9,
Works OK here, witht he latest of everything. Please send the faulty
.config.
If you could generate the `strace -f' output from good and bad
sessions and identify where things went wrong, that would help.
On Thu, 2004-10-28 at 14:07 +0530, Rajsekar wrote:
> No, that is not the actual problem (even though it seems to solve it).
> The thing is, /dev/tty should be rw for everyone.
> What happens is (it happened to me and CONFIG_LEGACY_PTYS is `n' in my
> case), that ssh tries to open /dev/tty and fails and ends up trying to use
> ssh-askpass (some other way to ask for password).
>
> But I am still not sure how CONFIG_LEGACY_PTYS=n solves it.
>
> Change the permissions of /dev/tty to something like 0666 or if you have
> udev, edit the permissions.d/* files
>
> Please give me your feedback.
>
I think when udev creates the ptys nodes dynamically it corrupts the
Legacy ones already create. With CONFIG_LEGACY_PTYS turned off there
never created and udev is free to do what it wants. It could be the
complete reverse though.
By the way th settings in permissions.d are 660 for most things
including ptys. I'm running slackware.
Dan.
On Thu, 28 Oct 2004, Andrew Morton wrote:
> Date: Thu, 28 Oct 2004 02:29:42 -0700
> From: Andrew Morton <[email protected]>
> To: Jesper Juhl <[email protected]>
> Cc: [email protected], [email protected]
> Subject: Re: SSH and 2.6.9
>
> Jesper Juhl <[email protected]> wrote:
> >
> > Now I guess we just need for someone to find out why LEGACY_PTYS breaks
> > ssh (and other apps?) with kernels >= 2.6.9,
>
> Works OK here, witht he latest of everything. Please send the faulty
> .config.
>
> If you could generate the `strace -f' output from good and bad
> sessions and identify where things went wrong, that would help.
>
I have no problem here, and I can't reproduce it by enabling LEGACY_PTYS
either, so you'll have to get the .config and strace etc from Danny Brow.
--
Jesper Juhl
> > Jesper Juhl <[email protected]> wrote:
> > >
> > > Now I guess we just need for someone to find out why LEGACY_PTYS
> > > breaks ssh (and other apps?) with kernels >= 2.6.9,
> >
> > Works OK here, witht he latest of everything. Please send
> > the faulty .config.
> >
> > If you could generate the `strace -f' output from good and bad
> > sessions and identify where things went wrong, that would help.
> >
>
> I have no problem here, and I can't reproduce it by enabling
> LEGACY_PTYS either, so you'll have to get the .config and
> strace etc from Danny Brow.
Similar problem occured at my box, but it's not kernel 2.6.9 which breaks
SSH. At leat for me...
If you are running FC3T3, then try to remount /dev. Since some version of
udev obsoletes the dev package and afterwards no /dev/pts/* exists, since
dev is no longer there and udev 'overmounted' /dev...
Best,
Oliver
Slackware 10 is (was?) known to have a faulty udev.rules files
breaking /dev/tty with latest 2.6 kernels. I already reported this to
Patrick Volkerding but have no idea if its fixed because I am now a
Debian user *g*.
Regards,
/ismail
On Fri, 29 Oct 2004 12:36:19 +0200 (CEST), Jesper Juhl <[email protected]> wrote:
> On Thu, 28 Oct 2004, Andrew Morton wrote:
>
> > Date: Thu, 28 Oct 2004 02:29:42 -0700
> > From: Andrew Morton <[email protected]>
> > To: Jesper Juhl <[email protected]>
> > Cc: [email protected], [email protected]
> > Subject: Re: SSH and 2.6.9
> >
> > Jesper Juhl <[email protected]> wrote:
> > >
> > > Now I guess we just need for someone to find out why LEGACY_PTYS breaks
> > > ssh (and other apps?) with kernels >= 2.6.9,
> >
> > Works OK here, witht he latest of everything. Please send the faulty
> > .config.
> >
> > If you could generate the `strace -f' output from good and bad
> > sessions and identify where things went wrong, that would help.
> >
>
> I have no problem here, and I can't reproduce it by enabling LEGACY_PTYS
> either, so you'll have to get the .config and strace etc from Danny Brow.
>
> --
> Jesper Juhl
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
--
Time is what you make of it
I am talking only about the file /dev/tty (IMHO you mistook it to mean /dev/ttyXX)
The permission should be 666 not 660 (for /dev/tty as its owner is always
root).
Try changing the permission of /dev/tty to 666 (rw-rw-rw-)
--
Rajsekar
On Fri, 29 Oct 2004, ismail d?nmez wrote:
> Date: Fri, 29 Oct 2004 13:55:33 +0300
> From: ismail d?nmez <[email protected]>
> To: Jesper Juhl <[email protected]>
> Cc: Andrew Morton <[email protected]>, [email protected],
> [email protected]
> Subject: Re: SSH and 2.6.9
>
> Slackware 10 is (was?) known to have a faulty udev.rules files
> breaking /dev/tty with latest 2.6 kernels. I already reported this to
> Patrick Volkerding but have no idea if its fixed because I am now a
> Debian user *g*.
>
That sounds plausible, since I'm running Slackware myself, but I'm running
slackware-current, and one of the things currently in -current is an udev
update. That would explain why I don't see the problem while Danny does.
--
Jesper Juhl