On Thu, 2012-04-26 at 13:51 +0200, Jakub Hrozek wrote:
> I scanned the SSSD source code with the clang static analyzer and found
> a number of issues. Patches are attached -- I think that most of them
> are OK to just include in master. Patches #1, #13 and #16 may be
> something that we want to include sooner.
> 
> Because the clang analyzer is freely available in Fedora, I think it would
> make sense to run a scan at least before a release.
> 
> Developers can also run the clang tool themselves - I'll add info how
> into our Developers page.


Patch 0001: Ack

Patch 0002: Ack

Patch 0003: Ack

Patch 0004: Nack
There is no NULL-termination guarantee here. We only initialize pid_str
with the first element being NULL.

Patch 0005: Ack

Patch 0006: Ack

Patch 0007: Ack

Patch 0008: Ack

Patch 0009: Ack

Patch 0010: Nack
If we're reporting ret (which is an errno_t value) I'd prefer if the
DEBUG messages included the strerror() conversion, since it's much
easier to read.

Patch 0011: Ack

Patch 0012: Ack

Patch 0013: Ack

Patch 0014: Nack
This is not the only case in ipa_get_autofs_options(), but returning
here leaks tmp_ctx and anything still attached to it. Please fix all
such cases here.

Patch 0015: Ack.

Patch 0016: Ack. 

Patch 0017: Ack.

Patch 0018: Nack
A tevent_req *_send() function should only return NULL if the
tevent_req_create() fails. Otherwise, all other failures should be
reported with tevent_req_error() and tevent_req_post().


I've pushed the acked patches to master (that's all but 0004, 0010, 0014
and 0018).

Attachment: signature.asc
Description: This is a digitally signed message part

_______________________________________________
sssd-devel mailing list
[email protected]
https://fedorahosted.org/mailman/listinfo/sssd-devel

Reply via email to