Bug 721002

(In reply to comment #1)
> Description of problem:
> On boot I now receive this message straight away:
> 
> alg: skcipher: Failed to load transform for ecb-aes-aesni: -2
> 
> Version-Release number of selected component (if applicable):
> Linux 2.6.38.8-35.fc15.x86_64
> 
> How reproducible:
> boot from Linux 2.6.38.8-35.fc15.x86_64 kernel
> 
> Steps to Reproduce:
> 1. Upgrade to Linux 2.6.38.8-35.fc15.x86_64
> 2. Boot
> 
> Actual results:
> alg: skcipher: Failed to load transform for ecb-aes-aesni: -2
> 

And what does it do after that? Fail to finish booting?

I have the same thing happening to me on the same platform on the new Kernel.
(In reply to comment #1)
> (In reply to comment #1)
> > Description of problem:
> > On boot I now receive this message straight away:
> > 
> > alg: skcipher: Failed to load transform for ecb-aes-aesni: -2
> > 
> > Version-Release number of selected component (if applicable):
> > Linux 2.6.38.8-35.fc15.x86_64
> > 
> > How reproducible:
> > boot from Linux 2.6.38.8-35.fc15.x86_64 kernel
> > 
> > Steps to Reproduce:
> > 1. Upgrade to Linux 2.6.38.8-35.fc15.x86_64
> > 2. Boot
> > 
> > Actual results:
> > alg: skcipher: Failed to load transform for ecb-aes-aesni: -2
> > 
> 
> And what does it do after that? Fail to finish booting?

Chuck, after the error shows up Fedora 1(w/gnome 3) boots and I can see no *visible* problem but the error always shows up.

(In reply to comment #0)
> Description of problem:
> On boot I now receive this message straight away:
> 
> alg: skcipher: Failed to load transform for ecb-aes-aesni: -2
> 
> Version-Release number of selected component (if applicable):
> Linux 2.6.38.8-35.fc15.x86_64
> 
> How reproducible:
> boot from Linux 2.6.38.8-35.fc15.x86_64 kernel
> 
> Steps to Reproduce:
> 1. Upgrade to Linux 2.6.38.8-35.fc15.x86_64
> 2. Boot
> 
> Actual results:
> alg: skcipher: Failed to load transform for ecb-aes-aesni: -2
> 
> Expected results:
> no error
> 
> Additional info:

I have the same issue (error message at boot) after a kernel update to 2.6.38.8-35.fc15.x86_64

I too have the same problem but the computer hungs at the graphic nvidia screen during startup.

> 
> And what does it do after that? Fail to finish booting?

No it boots and all is well in the world :)

Just reporting this new bug that has popped up in transition to the new kernel. Not sure if it would cause issues booting for those people with encrypted disks - I don't have any so no problem for me apart from the error.

In my case it does fail finish booting. After several minutes I've switched off the computer an rstarted with the 2.6.38.8-32.fc15.x86_64 kernel. Also I've marked the 2.6.38.8-35.fc15.x86_64 version in order yo avoid beeing loaded again.

Seeing the same thing, just before plymouth takes over. In addition there is another related error just after it.

alg: skcipher: Failed to load transform for ecb-aes-aesni: -2
alg: No test for __gcm-aes-aesni (__driver-gcm-aes-aesni)

kernel-2.6.38.8-35.fc15.x86_64 on a ThinkPad T410

The system does continue to boot, and I have not noticed any further issues from this.

It seems, looking at bug 589390 that this may be fixed upstream in kernel 3.0

(In reply to comment #8)
> It seems, looking at bug 589390 that this may be fixed upstream in kernel 3.0

You might try the 2.6.40-4 kernel that is in updates-testing for F15.  That is 3.0 rebranded.

Just tried it, but I am still getting the exact same failure.

Same problem here with kernel 2.6.40-4.fc15.x86_64
Also mentioned in my /var/log/boot.log: 
Aug  4 12:15:40 kiowa kernel: [    0.963618] alg: skcipher: Failed to load transform for ecb-aes-aesni: -2

same here goes for both kernel :

2.6.38.8-35.fc15.x86_64
2.6.40-4.fc15.x86_64

am still using the old kernel it seems to work fine 2.6.38.8-32.fc15.x86_64

I'm getting the same error message on boot up and shutdown.

skcipher: Failed to load transform for ecb-aes-aesni: -2

Just updated to 2.6.40-4.f15.x86_64, and I get the same error.
After that everything appears to be working fine.

There appears to be some kind of race between the AES_NI_INTEL module and the crypto manager framework.  I've asked upstream about this.

FYI: This message also shows up with F16 Alpha TC1/RC1/RC2/RC3.

Same error message here with 2.6.38.8-35.fc15.x86_64 when upgraded from kernel 2.6.38.6-26-rc1.fc15.x86_64.

Although the system can boot, I can no longer connect to our corporate wireless network which uses WPA2-Enterprise with AES encryption; the NetworkManager keeps asking for my account information. Before the upgrade, the system with kernel 2.6.38.6-26 was able to connect to the wireless network; after the upgrade to 2.6.38.8-35, neither can connect.

Would this problem be related to this ecb-aes-aesni error?
Thanks!

Upstream has provided a patch and I've committed it to the kernel branches.  It should be in the next rebuild.

kernel-3.1.0-0.rc3.git0.0.fc16 has been submitted as an update for Fedora 16.
https://admin.fedoraproject.org/updates/kernel-3.1.0-0.rc3.git0.0.fc16

Package kernel-3.1.0-0.rc3.git0.0.fc16:
* should fix your issue,
* was pushed to the Fedora 16 testing repository,
* should be available at your local mirror within two days.
Update it with:
# su -c 'yum update --enablerepo=updates-testing kernel-3.1.0-0.rc3.git0.0.fc16'
as soon as you are able to, then reboot.
Please go to the following url:
https://admin.fedoraproject.org/updates/kernel-3.1.0-0.rc3.git0.0.fc16
then log in and leave karma (feedback).

I also have an error 
"alg: skcipher: Failed to load transform for ecb-aes-aesni: -2" , after update, and system doesn't boot.
In the rescue mode I have installed kernel-3.1.0-0.rc3.git0.0.fc16, but it doesn't work too with similar error:
"xhci-ncd 000:02:00.0 Failed to enable MSI-X" 

In the result I've switched the kernel to 2.6.38 version.


In the process of setting up my server I also tried Fedora16-Alfa.
In the process of setup on some stage has been shown an error, like with 2.6.40 kernel, but the installation was going on.
(Also one note for someone who interested about Fedora16-alpha: I had problem with RAID on Samsung disks.
After of selection of RAID1 in setup process, installator said somewhat like: "disk has raid information but it doesn't participate in the RAID" and installation had failed. Fedora15 uses this RAID without problems)

(In reply to comment #21)
> I also have an error 
> "alg: skcipher: Failed to load transform for ecb-aes-aesni: -2" , after update,
> and system doesn't boot.

I'm guessing this was with a kernel that didn't have the fix.

> In the rescue mode I have installed kernel-3.1.0-0.rc3.git0.0.fc16, but it
> doesn't work too with similar error:
> "xhci-ncd 000:02:00.0 Failed to enable MSI-X" 

That error isn't similar at all.  It is entirely unrelated to this bug.

kernel-3.1.0-0.rc3.git0.0.fc16 has been pushed to the Fedora 16 stable repository.  If problems still persist, please make note of it in this bug report.

I repeated the update procedure from  Fedora 16 stable repository.
kernel-3.1.0-0.rc3.git0.0.fc16 has an error such as I reported before: "xhci-ncd 000:02:00.0 Failed to enable MSI-X" right after when messages were available on the screen.
First time my system was loaded (with error message), but then, on next loadings, was stopped, with message in the end:
"Dropping to debug shell
sh: Can't access tty; job control turned off"

I mentioned that this bug is similar because booting of systems in two cases was stopped on the same stage (or like stage).

Please do something, since Fedora15 has another bug https://bugzilla.redhat.com/show_bug.cgi?id=729340 , which was fixed in the kernel-2.6.40.3-0.fc15 and which(kernel)I can't use too.
(Excuse me for that proposal, but perhaps it is possible to make kernel-2.6.39 which will be just kernel-2.6.38 with fixed bug 729340, if work about kernel-3.1.0 will take too many time)

Vadim, please open a new bug for your issues.  They are unrelated to this bug.

Also, 729340 is fixed in 3.1-rc3 already.

I'm getting this issue on a Lenovo w520. Interestingly, the system boots fine with integrated graphics, but hangs at the alg: skcipher: Failed to load transform for ecb-aes-aesni: -2 error with discrete graphics. 

Unfortunately, discrete graphics appear to be required for external monitor support.

Am I correct in understanding that the only fix for this issue is upgrading to F16? I've not been able to successfully install F16 alpha on this machine. :(

(In reply to comment #26)
> Am I correct in understanding that the only fix for this issue is upgrading to
> F16? I've not been able to successfully install F16 alpha on this machine. :(

No, 2.6.40.3-2.fc15 or newer on f15 should have the fix as well.

Thanks, Josh. I just grabbed 2.6.40-4 from http://koji.fedoraproject.org/koji/buildinfo?buildID=256252, and I'm getting the same behavior.

(In reply to comment #28)
> Thanks, Josh. I just grabbed 2.6.40-4 from
> http://koji.fedoraproject.org/koji/buildinfo?buildID=256252, and I'm getting
> the same behavior.

That's because it's not 2.6.40.3-2.fc15 or newer.  That extra .3 is important.  Just use this one:

http://koji.fedoraproject.org/koji/buildinfo?buildID=261186

> 
> That's because it's not 2.6.40.3-2.fc15 or newer.  That extra .3 is important.

Thanks! I wondered whether I was reading my dots and dashes wrong. The error message is gone, but it look like my discrete graphics booting bug was unrelated, because I'm still getting the hang.

I'll see if it's been reported elsewhere, and file a bug if not.

(In reply to comment #30)
> > 
> > That's because it's not 2.6.40.3-2.fc15 or newer.  That extra .3 is important.
> 
> Thanks! I wondered whether I was reading my dots and dashes wrong. The error
> message is gone, but it look like my discrete graphics booting bug was
> unrelated, because I'm still getting the hang.
> 
> I'll see if it's been reported elsewhere, and file a bug if not.

Hi buddy,

Yeah the hang is unrelated to this bug - the discreet graphics error is what is causing your problem and would be related to your video card drivers. Gnome Shell is pretty particular about having working video drivers although it should have given you an option to fall back to gnome fallback which isn't as particular...

Do a search for your particular laptop and video drivers for that and you should be able to muddle through it

kernel-2.6.40.4-5.fc15 has been submitted as an update for Fedora 15.
https://admin.fedoraproject.org/updates/kernel-2.6.40.4-5.fc15

kernel-2.6.40.4-5.fc15 doesn't work on my system.

It has booted first time after the installation of 2.6.40.4-5 , but next times the loading of system was failed with no any messages on the screen, except for  final message:
"Dropping to debug shell
sh: Can't access tty; job control turned off"

First time (when the system was loaded successfully) during the boot I saw the message on the screen: 
"xhci-ncd 000:02:00.0 Failed to enable MSI-X" among other messages (not on the top), but this message has not been written down in the boot.log.
The message "Failed to enable MSI-X" is also on kernel  2.6.38 , but this kernel works.

(In reply to comment #33)
> kernel-2.6.40.4-5.fc15 doesn't work on my system.
> 
> It has booted first time after the installation of 2.6.40.4-5 , but next times
> the loading of system was failed with no any messages on the screen, except for
>  final message:
> "Dropping to debug shell
> sh: Can't access tty; job control turned off"
> 
> First time (when the system was loaded successfully) during the boot I saw the
> message on the screen: 
> "xhci-ncd 000:02:00.0 Failed to enable MSI-X" among other messages (not on the
> top), but this message has not been written down in the boot.log.
> The message "Failed to enable MSI-X" is also on kernel  2.6.38 , but this
> kernel works.

That is entirely unrelated to this bug.  Please open a new one as I said in comment #25

Statements about "MSI-X" just a note.
Mostly I wanted to report, that kernel-2.6.40.4-5.fc15 doesn't work at all.

kernel-2.6.40.4-5.fc15 has been pushed to the Fedora 15 stable repository.  If problems still persist, please make note of it in this bug report.

Painfully to speak, but this kernel doesn't work on my system (Asrock H67M-GE,I5-2400S,fake RAID1 on too HD154UI)
Message on the screen:
"
Dropping to debug shell
sh: Can't access tty; job control turned off
dracut:/#
"

Unfortunately I am still getting this error as well:
[fotios@fotios ~]$ sudo grep aes /var/log/messages
Sep 13 07:54:02 fotios kernel: [    0.801907] alg: skcipher: Failed to load transform for ecb-aes-aesni: -2
Sep 13 07:54:02 fotios kernel: [    0.802065] alg: No test for __gcm-aes-aesni (__driver-gcm-aes-aesni)
Sep 16 10:17:51 fotios kernel: [    0.792011] alg: No test for __gcm-aes-aesni (__driver-gcm-aes-aesni)
Sep 16 10:47:09 fotios kernel: [    0.792229] alg: No test for __gcm-aes-aesni (__driver-gcm-aes-aesni)
[fotios@fotios ~]$ uname -a
Linux fotios.redhat.com 2.6.40.4-5.fc15.x86_64 #1 SMP Tue Aug 30 14:38:32 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux

I am using a Lenovo T520

(In reply to comment #38)
> Unfortunately I am still getting this error as well:
> [fotios@fotios ~]$ sudo grep aes /var/log/messages
> Sep 13 07:54:02 fotios kernel: [    0.801907] alg: skcipher: Failed to load
> transform for ecb-aes-aesni: -2
> Sep 13 07:54:02 fotios kernel: [    0.802065] alg: No test for __gcm-aes-aesni
> (__driver-gcm-aes-aesni)

There's a 3 day gap in your logs here.  Did you upgrade and reboot?

> Sep 16 10:17:51 fotios kernel: [    0.792011] alg: No test for __gcm-aes-aesni
> (__driver-gcm-aes-aesni)
> Sep 16 10:47:09 fotios kernel: [    0.792229] alg: No test for __gcm-aes-aesni
> (__driver-gcm-aes-aesni)

Note that the first message "alg: skcipher: Failed to load transform for ecb-aes-aesni: -2" doesn't show up in the Sep 16 version.  That is what this bug fixed.  The "No test for" aren't errors.

You're right, my mistake. Thank you.

I still get this error when using fips=1

This is with 3.0.0-8 on an Lenovo X201 with intel aesni

How to repeat:

yum install dracut-fips dracut-fips-aesni

This is with dracut.conf modified with:

add_dracutmodules="masterkey integrity securityfs crypt-gpg fips fips-aesni"

and the fixes from Steve Grub at https://bugzilla.redhat.com/show_bug.cgi?id=808744

(In reply to comment #41)
> I still get this error when using fips=1
> 
> This is with 3.0.0-8 on an Lenovo X201 with intel aesni
> 
> How to repeat:
> 
> yum install dracut-fips dracut-fips-aesni
> 
> This is with dracut.conf modified with:
> 
> add_dracutmodules="masterkey integrity securityfs crypt-gpg fips fips-aesni"
> 
> and the fixes from Steve Grub at
> https://bugzilla.redhat.com/show_bug.cgi?id=808744

Paul, I'd prefer we track this in a new bug.  This bug has been closed for a while and is correctly closed out as fixed.  The issue you mention is only present with fips mode afaik.

There was a discussion here pointing to this bug:

https://lkml.org/lkml/2011/8/9/338

Seeing a similar thing with kernel-3.6.3-1.fc17.x86_64

alg: No test for __gcm-aes-aesni (__driver-gcm-aes-aesni)

Is there a new bug somewhere?