linux-kernel - Re: [RESEND PATCH] gpio_addr

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] [day] [month] [year] [list]

Date:	Tue, 17 Nov 2009 18:21:58 -0500
From:	Mike Frysinger <vapier.adi@...il.com>
To:	Alan Cox <alan@...rguk.ukuu.org.uk>
Cc:	Alan Cox <alan@...ux.intel.com>,
	Linux kernel mailing list <linux-kernel@...r.kernel.org>
Subject: Re: [RESEND PATCH] gpio_addr_flash: Fix warnings

On Tue, Nov 17, 2009 at 11:58, Alan Cox wrote:
> On Tue, 17 Nov 2009 11:50:24 -0500 Mike Frysinger wrote:
>> On Tue, Nov 17, 2009 at 10:46, Alan Cox wrote:
>> > --- a/drivers/mtd/maps/gpio-addr-flash.c
>> > +++ b/drivers/mtd/maps/gpio-addr-flash.c
>> > -       state->gpio_addrs     = (void *)gpios->start;
>> > +       state->gpio_addrs     = (void *)(unsigned long)gpios->start;
>> > -       state->map.virt       = (void __iomem *)memory->start;
>> > +       state->map.virt       = (void __iomem *)(unsigned long)memory->start;
>>
>> on what system do you see warnings ?  start should be resource_size_t
>> which should be phys_addr_t which should cast fine to a void*.
>
> 32bit x86 shows it up because resource_size_t is a physaddr_t which is
> 64bit (because of PAE). Now given you don't seem likely to be using the
> top bits for any real world example that looks ok but really casting a
> physical address to a pointer type needs care and the compiler can't
> deduce your intent there.

i dont think the casts will work in either case here and it's a good
thing that the code doesnt compile warning free

the gpio_addrs member is an array of gpio ids which are 32bit unsigned
values.  overlaying an array of 64bit values isnt going to produce
correct behavior.

as for the memory->start, i do need to access that directly, so that
should probably be going through ioremap_nocache() instead.

i.e. i'll have to write/test/post a patch rather than adding casts to
pretend there isnt a problem
-mike
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/