Make sure tools never create config.ldb #7551
Conversation
alexey-tikhonov
added
no-backport
alexey-tikhonov
force-pushed
the
confdb-init
branch
from
ea1b0d4
to
9a2c874
Compare
Take a note this also enforces check for existance of config.ldb in additional code paths (this is intentional side effect, see a008acc)
alexey-tikhonov
force-pushed
the
confdb-init
branch
2 times, most recently
from
8289db3
to
b94f8cf
Compare
Take a note this also enforces check for existance of config.ldb in additional code paths (this is intentional side effect, see a008acc)
alexey-tikhonov
force-pushed
the
confdb-init
branch
from
b94f8cf
to
12d38cc
Compare
| @@ -108,18 +112,14 @@ static errno_t sss_tool_confdb_init(TALLOC_CTX *mem_ctx, | |||
| return ret; | |||
| } | |||
|
|
|||
| ret = confdb_init(mem_ctx, &confdb, path); | |||
| ret = confdb_init(mem_ctx, _confdb, path); | |||
There was a problem hiding this comment.
Not directly related to you commit.
confdb_init() is assuming its cdb_ctx parameter is not NULL. It clearly is not here, but could be in other calls. Shouldn't we also check inside this function that it is not NULL? Or maybe just move the check to that function?
There was a problem hiding this comment.
Shouldn't we also check inside this function that it is not NULL?
this == sss_tool_confdb_init()?
But here is the check in place at the beginning (line 97)
Or do I misunderstand?
There was a problem hiding this comment.
No, sorry. this == confdb_init().
In other words, sss_tool_confdb_init() is checking _confdb != NULL, but other functions are also calling confdb_init(), and those functions may or may not check that condition. It might be better to move the check inside confdb_init().
There was a problem hiding this comment.
Ok, added one more patch.
closer to a place where argument is really used
alexey-tikhonov
added
Accepted
Ready to push
This is an addition to a008acc