On Mon, Aug 03, 2015 at 02:33:27PM +0100, Christoph Hellwig wrote: > On Mon, Aug 03, 2015 at 11:43:09AM +0100, Catalin Marinas wrote: > > The simplest would be to add a kmemleak_not_leak() annotation in > > scsi_init_request(), though you would hide real leaks (if any). > > > > A better way could be to inform kmemleak of these pages, something like > > below (compile-tested only): > > Both versions sound reasonable to me, but I'd prefer the second one if > it works.
I don't currently have a system where scsi_init_request() is used. But I hacked loop_init_request (and added a loop_exit_request) to allocate a dummy structure. Without this patch, I indeed get a few hundred kmemleak reports which disappear once I apply it. -- Catalin -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
