linux-kernel - Re: [PATCH V2] staging: fieldbus: anybus-s: force endiannes annotation

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: <20190430041934.GI23075@ZenIV.linux.org.uk>
Date:   Tue, 30 Apr 2019 05:19:34 +0100
From:   Al Viro <viro@...iv.linux.org.uk>
To:     Nicholas Mc Guire <der.herr@...r.at>
Cc:     Sven Van Asbroeck <thesven73@...il.com>,
        Nicholas Mc Guire <hofrat@...dl.org>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        devel@...verdev.osuosl.org,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH V2] staging: fieldbus: anybus-s: force endiannes
 annotation

On Tue, Apr 30, 2019 at 05:33:10AM +0200, Nicholas Mc Guire wrote:

> ok - my bad thn - I had assumed that using __force is reasonable
> if the handling is correct and its a localized conversoin only 
> like var = be16_to_cpu(var) which evaded introducing additinal
> variables just to have different types but no different function.

If compiler can't recognize that in

	T1 v1;
	T2 v2;

	code using v1, but not v2
	v2 = f(v1);
	code using v2, but not v1

it can use the same memory for v1 and v2, file a bug against the
compiler.  Or stop using that toy altogether - that kind of
optimizations is early 60s stuff and any real compiler will
handle that.  Both gcc and clang certainly do handle that.

Another thing they handle is figuring out that be16_to_cpu()
et.al. are pure functions, so

	f(be16_to_cpu(n));
	no modifications of n
	g(be16_to_cpu(n));

doesn't need to have le16_to_cpu recalculated.  IOW, that particular
code could as well have been
	dev_info(dev, "Fieldbus type: %04X", be16_to_cpu(fieldbus_type));
	...
	cd->client->fieldbus_type = be16_to_cpu(fieldbus_type);

... not that there's much sense keeping ->fieldbus_type in host-endian,
while we are at it.