Ben, Thanks for reviewing, On 12/12/2016 12:23 PM, Benjamin Marzinski wrote:
Why don't we just call pathinfo here? I see that you set the pathvec to NULL so that you don't actually store the path, but this sure does a lot of unnecessary work before failing. AFAICT, just calling pathinfo with DI_BLACKLIST already gets you what you need (it checks both the device and wwid blacklists). Am I missing something here?
path_discover() also calls filter_property() and filter_devnode(), so more blacklisting methods are covered. (filter_devnode() is also checked by uev_trigger(), but I wanted to reuse the existing code to keep the patch simple.) And afaik there is not a lot more being done in path_discover() > store_pathinfo() > path_info() vs the direct path_info() call; There /is/ some overhead, you're right, but since this is only in the error path of this corner case, it seemed to be not a big deal. -- Mauricio Faria de Oliveira IBM Linux Technology Center -- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
