linux-kernel - Re: [RFC PATCH 2/5] lib/memregion: Uplevel the pmem "region" ida to a global allocator

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]

Date:   Thu, 4 Apr 2019 12:32:11 -0700
From:   Matthew Wilcox <willy@...radead.org>
To:     Dan Williams <dan.j.williams@...el.com>
Cc:     linux-kernel@...r.kernel.org, Keith Busch <keith.busch@...el.com>,
        vishal.l.verma@...el.com, x86@...nel.org, linux-mm@...ck.org,
        linux-nvdimm@...ts.01.org
Subject: Re: [RFC PATCH 2/5] lib/memregion: Uplevel the pmem "region" ida to
 a global allocator

On Thu, Apr 04, 2019 at 12:08:38PM -0700, Dan Williams wrote:
> +++ b/lib/Kconfig
> @@ -318,6 +318,12 @@ config DECOMPRESS_LZ4
>  config GENERIC_ALLOCATOR
>  	bool
>  
> +#
> +# Generic IDA for memory regions
> +#

Leaky abstraction -- nobody needs know that it's implemented as an IDA.
Suggest:

# Memory region ID allocation

...

> +++ b/lib/memregion.c
> @@ -0,0 +1,22 @@
> +#include <linux/idr.h>
> +#include <linux/module.h>
> +
> +static DEFINE_IDA(region_ida);
> +
> +int memregion_alloc(void)
> +{
> +	return ida_simple_get(&region_ida, 0, 0, GFP_KERNEL);
> +}
> +EXPORT_SYMBOL(memregion_alloc);
> +
> +void memregion_free(int id)
> +{
> +	ida_simple_remove(&region_ida, id);
> +}
> +EXPORT_SYMBOL(memregion_free);
> +
> +static void __exit memregion_exit(void)
> +{
> +	ida_destroy(&region_ida);
> +}
> +module_exit(memregion_exit);

 - Should these be EXPORT_SYMBOL_GPL?
 - Can we use the new interface, ida_alloc() and ida_free()?
 - Do we really want memregion_exit() to happen while there are still IDs
   allocated in the IDA?  I think this might well be better as:

	BUG_ON(!ida_empty(&region_ida));

Also, do we really want to call the structure the region_ida?  Why not
region_ids?