| commit | e54409ef18a88a88553f7a705cd6c04cda9ca238 | |
| author | Dan Carpenter <dan.carpenter@oracle.com> | |
| Thu, 28 Jun 2018 07:24:39 +0000 (28 10:24 +0300) | ||
| committer | Dan Carpenter <dan.carpenter@oracle.com> | |
| Thu, 28 Jun 2018 07:25:12 +0000 (28 10:25 +0300) | ||
| tree | a0c27e0fa06f61de45b1c93cd0b1deff3caa3746 | treesnapshot (tar.gz zip) |
| parent | e2a33714b1415a3c729173e9de4a0de893907b0a | commitdiff |
nospec: if it's called with trusted data that's the same as nospec
The problem here is that when we mix user, nospec data with trusted data,
the result should be nospec.
I made this only true when we're dealing with caller_info. It should
probably just be true all the time. I think it doesn't matter and this
felt cautious and conservative to me when I wrote it.
Reported-by: Josh Poimboeuf <jpoimboe@redhat.com>
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
The problem here is that when we mix user, nospec data with trusted data,
the result should be nospec.
I made this only true when we're dealing with caller_info. It should
probably just be true all the time. I think it doesn't matter and this
felt cautious and conservative to me when I wrote it.
Reported-by: Josh Poimboeuf <jpoimboe@redhat.com>
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
| check_nospec.c | diffblobblamehistory |