[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

[libvirt] syntax-check vs. close [was: [PATCH 1/3] Cancel migration if user presses Ctrl-C when migration is in progress]

On 01/25/2011 12:05 AM, Wen Congyang wrote:
>>> +cleanup:
>>> >> +    if (dom)
>>> >> +        virDomainFree(dom);

My bad for thinking this was a violation of the useless-if-before-free
rule; virDomainFree is _not_ free-like: it returns an int, and issues an
error if domain is NULL.

>>> >> +    if (p[0] != -1) {
>>> >> +        close(p[0]);
>>> >> +        close(p[1]);
>> > 
>> > 'make syntax-check' should have caught these violations (if it didn't,
>> > then that's something that we should fix).  This should be:
> unfortunately, 'make syntax-check' does not catch these violations.

Oops. Stupid typo on my part, back in commit f1fe9671.  Full patch
coming up, once I fix the remaining violations that the typo fix caught.
 Oh, and I should probably add a rule to prohibit popen while I'm at it
(virCommand is better, but I thought of it because pclose is problematic).

diff --git i/cfg.mk w/cfg.mk
index 066fa3d..10eaab2 100644
--- i/cfg.mk
+++ w/cfg.mk
@@ -241,7 +241,7 @@ sc_avoid_write:

 # Avoid functions that can lead to double-close bugs.
-	@prohibit='\<[f]close *\('					\
+	@prohibit='\<[f]?close *\('					\
 	halt='use VIR_{FORCE_}[F]CLOSE instead of [f]close'		\
 	@prohibit='\<fdopen *\('					\

Eric Blake   eblake redhat com    +1-801-349-2682
Libvirt virtualization library http://libvirt.org

Attachment: signature.asc
Description: OpenPGP digital signature

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]