On Thu, 8 Aug 2013, Joe Perches wrote: > On Thu, 2013-08-08 at 21:32 +0200, Julia Lawall wrote: > > These patches result from the following semantic patch > > (http://coccinelle.lip6.fr/), which checks for expression statements > > separated by , rather than by ;. The rule is quite conservative in that it > > only finds cases where this pattern appears after a statement that ends > > with ;, to ensure that the expression statements do not form an if branch > > or loop body. Iteration is used to treat arbitrary sequences of commas. > > Nice. Thanks. > Maybe there also some cases where a block close brace > is followed by a statement , then statement ; > > Like from fs/reiserfs/fix_node.c > > 231 if ((unsigned int)cur_free >= > 232 (vn->vn_size - > 233 ((vi->vi_type & VI_TYPE_LEFT_MERGEABLE) ? IH_SIZE : 0))) { > 234 /* all contents of S[0] fits into L[0] */ > 235 > 236 RFALSE(vn->vn_mode == M_INSERT || vn->vn_mode == M_PASTE, > 237 "vs-8055: invalid mode or balance condition failed"); > 238 > 239 tb->lnum[0] = vn->vn_nr_item; > 240 tb->lbytes = -1; > 241 return; > 242 } > 243 > 244 d_size = 0, ih_size = IH_SIZE; I did think of a } as another safe case, but actually, I would not have done anything with this one because they are both on the same line. But I was more thinking of i++, j++; which I saw somewhere. This one is a bit more complicated. and would probably be better on separate lines. Thanks for the feedback. julia -- To unsubscribe from this list: send the line "unsubscribe kernel-janitors" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
