Bug 182598 - ia64: /dev/efirtc accessed before udev runs
Summary: ia64: /dev/efirtc accessed before udev runs
Keywords:
Status: CLOSED NEXTRELEASE
Alias: None
Product: Fedora
Classification: Fedora
Component: mkinitrd
Version: rawhide
Hardware: All
OS: Linux
medium
medium
Target Milestone: ---
Assignee: Peter Jones
QA Contact: David Lawrence
URL:
Whiteboard:
Depends On: 182597
Blocks: fedora-ia64
TreeView+ depends on / blocked
 
Reported: 2006-02-23 16:18 UTC by Prarit Bhargava
Modified: 2007-11-30 22:11 UTC (History)
3 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2006-08-10 20:05:34 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
Additional patch for mkinitrd to correctly initialize /dev/efirtc (329 bytes, patch)
2006-05-17 20:10 UTC, Prarit Bhargava
no flags Details | Diff

Description Prarit Bhargava 2006-02-23 16:18:20 UTC
Description of problem:  
  
On ia64 platforms, hwclock is accessed via /dev/efirtc.  This file is 
'created' by udev which runs AFTER the hwclock sync is performed. 
  
Version-Release number of selected component (if applicable):  latest rawhide  
pull (fedora-devel branch)  
  
How reproducible: 100%  
  
  
Steps to Reproduce:  
1. Install Fedora ia64  
2. Boot.  
3.  
    
Actual results:  
  
Following error is seen on the console:  
  
Cannot access the Hardware Clock via any known method.  
Use the --debug option to see the details of our search for an access method.  
Setting clock  (localtime): Tue Feb 21 20:00:15 EST 2006 [  OK  ]  
Starting udev   
  
Expected results:  
  
No errors should be seen.  
  
Additional info:  
 
The bug is that /dev/efirtc does not exist. 
 
This differs from RHEL -- in that case the clock starts AFTER udev has run and 
the modules have loaded: 
 
                Press 'I' to enter interactive startup. 
Starting udev:  [  OK  ] 
Initializing hardware...  storage network audio done[  OK  ] 
Configuring kernel parameters:  [  OK  ] 
Setting clock  (localtime): Tue Feb 21 11:07:55 EST 2006 [  OK  ] 
 
Is there a reason that the rc.sysinit has been modified to startup the clock 
before udev and module loading?

Comment 1 Bill Nottingham 2006-02-23 16:36:42 UTC
/dev/rtc is handled in mkinitrd. /dev/efirtc probably should be as well.


Comment 2 Peter Jones 2006-02-23 23:22:33 UTC
So what's the correct major/minor for the device?

Comment 3 Prarit Bhargava 2006-02-24 01:41:06 UTC
Peter: 
 
[root@altix3 ~]# ls -l /dev/efirtc  
crw------- 1 root root 10, 136 Feb 22 08:31 /dev/efirtc 
[root@altix3 ~]#  
 
P. 

Comment 4 Peter Jones 2006-02-24 04:41:30 UTC
Ok, I've committed it to cvs head.  Next build will have it.

Comment 5 Prarit Bhargava 2006-05-17 20:10:02 UTC
Created attachment 129358 [details]
Additional patch for mkinitrd to correctly initialize /dev/efirtc 

Peter,

This patch is needed to fix the problem ...

P.

Comment 6 Prarit Bhargava 2006-08-10 20:05:34 UTC
Hmmm ... the patch above seems to have a bug in it :(

Comment 7 Keiichiro Tokunaga 2006-08-21 15:56:43 UTC
Prarit,

Fujitsu has opened an IT related to this problem (IT100192) and
Mori-san gave us a pointer to this bug.

I tested the patch above on my box (PRIMEQUEST) and it seems that
it fixed the problem.  According to your last comments, there may
still be a bug in the patch.  Could you tell me what kind of bug
do you think there is in the patch?  I'd like to share that information
with Fujitsu engineers if appropriate.

Comment 8 Prarit Bhargava 2006-08-21 16:24:26 UTC
Kei,

This BZ is closed, however, please see 

https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=202110

P.

Comment 9 Keiichiro Tokunaga 2006-08-21 16:42:37 UTC
Prarit,

OK.  I understand the situation.

Thank!


Note You need to log in before you can comment on or make changes to this bug.