[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <6329f91293668_2a6ded2948c@dwillia2-xfh.jf.intel.com.notmuch>
Date: Tue, 20 Sep 2022 10:32:02 -0700
From: Dan Williams <dan.j.williams@...el.com>
To: Bo Liu <liubo03@...pur.com>, <dan.j.williams@...el.com>,
<vishal.l.verma@...el.com>, <dave.jiang@...el.com>,
<ira.weiny@...el.com>
CC: <nvdimm@...ts.linux.dev>, <linux-kernel@...r.kernel.org>,
Bo Liu <liubo03@...pur.com>
Subject: RE: [PATCH] nvdimm: Call ida_simple_remove() when failed
Bo Liu wrote:
> In function nvdimm_bus_register(), when code execution fails, we should
> call ida_simple_remove() to free ida.
>
> Signed-off-by: Bo Liu <liubo03@...pur.com>
> ---
> drivers/nvdimm/bus.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/drivers/nvdimm/bus.c b/drivers/nvdimm/bus.c
> index b38d0355b0ac..3415dc62632b 100644
> --- a/drivers/nvdimm/bus.c
> +++ b/drivers/nvdimm/bus.c
> @@ -371,6 +371,7 @@ struct nvdimm_bus *nvdimm_bus_register(struct device *parent,
> return nvdimm_bus;
> err:
> put_device(&nvdimm_bus->dev);
> + ida_simple_remove(&nd_ida, nvdimm_bus->id);
No, this is a double-free. The put_device() will trigger
nvdimm_bus_release().
Powered by blists - more mailing lists