2024-03-14 08:10:54

by Karel Balej

[permalink] [raw]
Subject: Re: [PATCH] Revert "crypto: pkcs7 - remove sha1 support"

Eric,

Eric Biggers, 2024-03-13T16:32:27-07:00:
> From: Eric Biggers <[email protected]>
>
> This reverts commit 16ab7cb5825fc3425c16ad2c6e53d827f382d7c6 because it
> broke iwd. iwd uses the KEYCTL_PKEY_* UAPIs via its dependency libell,
> and apparently it is relying on SHA-1 signature support. These UAPIs
> are fairly obscure, and their documentation does not mention which
> algorithms they support. iwd really should be using a properly
> supported userspace crypto library instead. Regardless, since something
> broke we have to revert the change.
>
> It may be possible that some parts of this commit can be reinstated
> without breaking iwd (e.g. probably the removal of MODULE_SIG_SHA1), but
> for now this just does a full revert to get things working again.
>
> Reported-by: Karel Balej <[email protected]>
> Closes: https://lore.kernel.org/r/[email protected]
> Cc: Dimitri John Ledkov <[email protected]>
> Signed-off-by: Eric Biggers <[email protected]>

thank you very much for the revert. I have compiled 6.8 with this patch
and attest that it solves my eduroam connection issue.

Tested-by: Karel Balej <[email protected]>

May I please ask, though, why you did not Cc stable (and add a Fixes
trailer for that matter)? It seems like something that would be nice to
see fixed in 6.7.y and 6.8.y too as soon as possible.

Kind regards,
K. B.


2024-03-14 19:40:43

by Eric Biggers

[permalink] [raw]
Subject: Re: [PATCH] Revert "crypto: pkcs7 - remove sha1 support"

On Thu, Mar 14, 2024 at 09:11:08AM +0100, Karel Balej wrote:
> Eric,
>
> Eric Biggers, 2024-03-13T16:32:27-07:00:
> > From: Eric Biggers <[email protected]>
> >
> > This reverts commit 16ab7cb5825fc3425c16ad2c6e53d827f382d7c6 because it
> > broke iwd. iwd uses the KEYCTL_PKEY_* UAPIs via its dependency libell,
> > and apparently it is relying on SHA-1 signature support. These UAPIs
> > are fairly obscure, and their documentation does not mention which
> > algorithms they support. iwd really should be using a properly
> > supported userspace crypto library instead. Regardless, since something
> > broke we have to revert the change.
> >
> > It may be possible that some parts of this commit can be reinstated
> > without breaking iwd (e.g. probably the removal of MODULE_SIG_SHA1), but
> > for now this just does a full revert to get things working again.
> >
> > Reported-by: Karel Balej <[email protected]>
> > Closes: https://lore.kernel.org/r/[email protected]
> > Cc: Dimitri John Ledkov <[email protected]>
> > Signed-off-by: Eric Biggers <[email protected]>
>
> thank you very much for the revert. I have compiled 6.8 with this patch
> and attest that it solves my eduroam connection issue.
>
> Tested-by: Karel Balej <[email protected]>
>
> May I please ask, though, why you did not Cc stable (and add a Fixes
> trailer for that matter)? It seems like something that would be nice to
> see fixed in 6.7.y and 6.8.y too as soon as possible.

I just forgot. Reverts usually get backported without asking anyway, but the
following should be added to make it explicit:

Fixes: 16ab7cb5825f ("crypto: pkcs7 - remove sha1 support")
Cc: [email protected]

That should just be added when the patch is applied, unless I happen to need to
send out a new version anyway.

We need to decide who is actually going to apply this revert. Probably Herbert,
since he took the commit that's being reverted?

- Eric