linux-kernel - Re: [PATCH v3 5/7] time: Add struct __kernel

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: <20180712082609.GB8802@infradead.org>
Date:   Thu, 12 Jul 2018 01:26:09 -0700
From:   Christoph Hellwig <hch@...radead.org>
To:     Deepa Dinamani <deepa.kernel@...il.com>
Cc:     tglx@...utronix.de, linux-kernel@...r.kernel.org, arnd@...db.de,
        y2038@...ts.linaro.org
Subject: Re: [PATCH v3 5/7] time: Add struct __kernel_timex

I don't think this patches makes sense without the next one,
which actually uses the structure.

> +/* CONFIG_64BIT_TIME enables new 64 bit time_t syscalls in the compat path
> + * and 32-bit emulation.
> + */

Wrong comment style, also the 'compat path is the 32 (or 31 in case of
s390) bit emulation, so the comment seems rather confusing.

> +#ifndef CONFIG_64BIT_TIME
> +#define __kernel_timex timex
> +#endif

using #defines for structs has all kinds of ill effects.  Why can't
we aways use __kernel_timex for the in-kernel usage?

> +#ifndef __kernel_timex
> +struct __kernel_timex {
> +	unsigned int modes;	/* mode selector */
> +	int :32;            /* pad */

Why do we need padding for a purely in-kernel structure?

Also the anonymous member syntax is rather odd and I don't remeber
us using it anywhere else.  Why here?