linux-kernel - Re: [PATCH] libnvdimm, region: Use struct

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <d940183a-c00d-3a96-37bb-9553583f160a@embeddedor.com>
Date:   Wed, 28 Aug 2019 14:36:59 -0500
From:   "Gustavo A. R. Silva" <gustavo@...eddedor.com>
To:     "Verma, Vishal L" <vishal.l.verma@...el.com>,
        "Williams, Dan J" <dan.j.williams@...el.com>,
        "Jiang, Dave" <dave.jiang@...el.com>,
        "Busch, Keith" <keith.busch@...el.com>,
        "Weiny, Ira" <ira.weiny@...el.com>
Cc:     "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "linux-nvdimm@...ts.01.org" <linux-nvdimm@...ts.01.org>
Subject: Re: [PATCH] libnvdimm, region: Use struct_size() in kzalloc()

Hi Vishal,

On 8/28/19 1:51 PM, Verma, Vishal L wrote:

[..]

> 
> Hi Gustavo,
> 
> The patch looks good to me, however it looks like it might've missed
> some instances where this replacement can be performed?
> 

struct_size() does not apply to those scenarios. See below...

> 
> [1]: https://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm.git/tree/drivers/nvdimm/region_devs.c#n1030

struct_size() only applies to structures of the following kind:

struct foo {
   int stuff;
   struct boo entry[];
};

and this scenario includes two different structures:

struct nd_region {

	...

        struct nd_mapping mapping[0];
};

struct nd_blk_region {

	...

        struct nd_region nd_region;
};

> [2]: https://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm.git/tree/drivers/nvdimm/region_devs.c#n96
> 

In this scenario struct_size() does not apply directly because of the following
logic before the call to devm_kzalloc():

 	size_t flush_data_size = sizeof(void *);
	
	[..]

        for (i = 0; i < nd_region->ndr_mappings; i++) {

		[..]

                /* at least one null hint slot per-dimm for the "no-hint" case */
                flush_data_size += sizeof(void *);
		
		[..]

                flush_data_size += nvdimm->num_flush * sizeof(void *);
        }

Thanks
--
Gustavo