Re: [PATCH] Introduce v3 namespaced file capabilities

2017-04-27 Thread Eric W. Biederman
"Serge E. Hallyn" writes: > Quoting Eric W. Biederman (ebied...@xmission.com): >> ebied...@xmission.com (Eric W. Biederman) writes: >> >> > "Serge E. Hallyn" writes: >> > >> >> Quoting Eric W. Biederman (ebied...@xmission.com): >> >>> >> >>> "Serge E. Hallyn" writes: >> >>> >> >>> > diff --g

Re: [PATCH] Introduce v3 namespaced file capabilities

2017-04-27 Thread Serge E. Hallyn
Quoting Eric W. Biederman (ebied...@xmission.com): > ebied...@xmission.com (Eric W. Biederman) writes: > > > "Serge E. Hallyn" writes: > > > >> Quoting Eric W. Biederman (ebied...@xmission.com): > >>> > >>> "Serge E. Hallyn" writes: > >>> > >>> > diff --git a/fs/xattr.c b/fs/xattr.c > >>> > in

Re: [PATCH] Introduce v3 namespaced file capabilities

2017-04-27 Thread Eric W. Biederman
ebied...@xmission.com (Eric W. Biederman) writes: > "Serge E. Hallyn" writes: > >> Quoting Eric W. Biederman (ebied...@xmission.com): >>> >>> "Serge E. Hallyn" writes: >>> >>> > diff --git a/fs/xattr.c b/fs/xattr.c >>> > index 7e3317c..75cc65a 100644 >>> > --- a/fs/xattr.c >>> > +++ b/fs/xattr

Re: [PATCH] Introduce v3 namespaced file capabilities

2017-04-22 Thread Eric W. Biederman
"Serge E. Hallyn" writes: > Quoting Eric W. Biederman (ebied...@xmission.com): >> >> "Serge E. Hallyn" writes: >> >> Overall this looks quite reasonable. >> >> My only big concern was the lack of verifying of magic_etc. As without > > Yes, I was relying too much on the size check. > >> that

Re: [PATCH] Introduce v3 namespaced file capabilities

2017-04-22 Thread Serge E. Hallyn
Quoting Eric W. Biederman (ebied...@xmission.com): > > "Serge E. Hallyn" writes: > > Overall this looks quite reasonable. > > My only big concern was the lack of verifying of magic_etc. As without Yes, I was relying too much on the size check. > that the code might not be future compatible w

Re: [PATCH] Introduce v3 namespaced file capabilities

2017-04-21 Thread Eric W. Biederman
"Serge E. Hallyn" writes: > Root in a non-initial user ns cannot be trusted to write a traditional > security.capability xattr. If it were allowed to do so, then any > unprivileged user on the host could map his own uid to root in a private > namespace, write the xattr, and execute the file wit

Re: [PATCH] Introduce v3 namespaced file capabilities

2017-04-21 Thread Eric W. Biederman
"Serge E. Hallyn" writes: > Quoting Eric W. Biederman (ebied...@xmission.com): >> >> Serge, >> >> Is there any change of a Signed-off-by on this patch? Otherwise I don't >> think we can merge it. > > For pete's sake! I'm sorry, i seem to remember with just about every > other project other th

Re: [PATCH] Introduce v3 namespaced file capabilities

2017-04-21 Thread Serge E. Hallyn
Quoting Eric W. Biederman (ebied...@xmission.com): > > Serge, > > Is there any change of a Signed-off-by on this patch? Otherwise I don't > think we can merge it. For pete's sake! I'm sorry, i seem to remember with just about every other project other than this. particular. patch. Does this

Re: [PATCH] Introduce v3 namespaced file capabilities

2017-04-21 Thread Eric W. Biederman
Serge, Is there any change of a Signed-off-by on this patch? Otherwise I don't think we can merge it. Eric "Serge E. Hallyn" writes: > Root in a non-initial user ns cannot be trusted to write a traditional > security.capability xattr. If it were allowed to do so, then any > unprivileged use

[PATCH] Introduce v3 namespaced file capabilities

2017-04-19 Thread Serge E. Hallyn
Root in a non-initial user ns cannot be trusted to write a traditional security.capability xattr. If it were allowed to do so, then any unprivileged user on the host could map his own uid to root in a private namespace, write the xattr, and execute the file with privilege on the host. However sup