Bug 51044

Summary:	login is (still) broken, console logins have no controlling tty
Product:	[Retired] Red Hat Public Beta	Reporter:	Bill Crawford <billc>
Component:	util-linux	Assignee:	Elliot Lee <sopwith>
Status:	CLOSED RAWHIDE	QA Contact:	David Lawrence <dkl>
Severity:	high	Docs Contact:
Priority:	high
Version:	roswell
Target Milestone:	---
Target Release:	---
Hardware:	i386
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2001-08-23 13:59:14 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Bill Crawford 2001-08-06 21:17:58 UTC

Description of Problem:
^C, ^\ etc. fail to work on console login.
vlock fails with "vlock: could not open /dev/tty: No such device or
address"

How Reproducible:
Log in at console, and try any of the above :o)

Steps to Reproduce:
1. Log in at the console
2. Run any command e.g. make, then try to interrupt it, or
3. try to run "vlock"

Actual Results:
Pressing ^C fails to interrupt make; vlock fails with the above error
message.

Expected Results:
Make should quit with an "interrupted" message; vlock should lock the
console, vlock -a should also prevent VT switching.

Additional Information:
Appears to be related to code in login to do a "TIOCNOTTY" ioctl in the
parent process.

Comment 1 Glen Foster 2001-08-06 22:53:15 UTC

We (Red Hat) really need to fix this defect before next release.

Comment 2 Bill Crawford 2001-08-06 22:59:34 UTC

*cough*

Yes, but I think it needs feeding back to the util-linux maintainer, Andries
Brouwer <aeb>.

I could do that, but I suspect you guys know more about the cause of / fix for
this than I do.

Comment 3 Elliot Lee 2001-08-07 17:29:13 UTC

glen wasn't making any comment about feeding patches upstream, just making sure
someone does something about the bug.

I can't reproduce the problem. mingetty-0.9.4-18, SysVinit-2.78-18,
util-linux-2.11f-8.

Can you tell me what 'ls -l /dev/tty `tty`' looks like on one of these logins?
Also the output of 'stty' and 'echo $TERM'? Thanks.

Comment 4 Bill Crawford 2001-08-07 17:53:42 UTC

Interesting results:

[bill@billc bill]$ ls -l /dev/tty `tty`
crw-rw-rw-    1 root     root       5,   0 Jul 18 16:58 /dev/tty
crw--w----    1 bill     tty        4,   2 Aug  7 18:53 /dev/tty2
[bill@billc bill]$ tty
/dev/tty2
[bill@billc bill]$ vlock
vlock: could not open /dev/tty: No such device or address

It was the vlock problem that originally alerted me.  The ^C issue is the
symptom that was originally reported to you.

As far as package versions go, I upgraded on top of Raw Hide from July 13th, but
seem to have roughly the same:

[bill@billc bill]$ rpm -q mingetty SysVinit util-linux
mingetty-0.9.4-18
SysVinit-2.78-18
util-linux-2.11f-5

I guess it may have been fixed between the -5 and -8 releases :o)

Is it possible to update Raw Hide a bit faster at times like this?

Comment 5 Bill Crawford 2001-08-07 17:56:36 UTC

Sorry, missed a piece:

[bill@billc bill]$ stty
speed 38400 baud; line = 0;
-brkint -imaxbel
[bill@billc bill]$ echo $TERM
linux

But it sounds like your internal version may have solved the problem by now. 
Does "vlock" work for you when you log in to a console?

Comment 6 Elliot Lee 2001-08-07 18:13:43 UTC

Try updating your util-linux for sure, I'll try getting a rawhide push RSN.

Comment 7 Trond Eivind Glomsrxd 2001-08-21 14:33:14 UTC

*** Bug 51212 has been marked as a duplicate of this bug. ***

Comment 8 Bill Crawford 2001-08-21 18:50:14 UTC

Seems to be fixed in the latest Raw Hide push, actually.
As far as I'm concerned this can be closed now.

Comment 9 Nalin Dahyabhai 2001-08-23 13:59:10 UTC

*** Bug 52051 has been marked as a duplicate of this bug. ***

Comment 10 Nalin Dahyabhai 2001-08-23 14:00:04 UTC

Appears to be fixed here as well.  Closing.