[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <alpine.DEB.2.20.1605171538040.24255@digraph.polyomino.org.uk>
Date: Tue, 17 May 2016 15:40:47 +0000
From: Joseph Myers <joseph@...esourcery.com>
To: Szabolcs Nagy <szabolcs.nagy@....com>
CC: Yury Norov <ynorov@...iumnetworks.com>, <arnd@...db.de>,
<catalin.marinas@....com>, <linux-arm-kernel@...ts.infradead.org>,
<linux-kernel@...r.kernel.org>, <nd@....com>,
<linux-arch@...r.kernel.org>, <linux-s390@...r.kernel.org>,
<pinskia@...il.com>, <Prasun.Kapoor@...iumnetworks.com>,
<schwab@...e.de>, <broonie@...nel.org>,
<linux-doc@...r.kernel.org>, <heiko.carstens@...ibm.com>,
<agraf@...e.de>, <klimov.linux@...il.com>,
<bamvor.zhangjian@...wei.com>, <schwidefsky@...ibm.com>,
<Nathan_Lynch@...tor.com>,
<christoph.muellner@...obroma-systems.com>,
GNU C Library <libc-alpha@...rceware.org>
Subject: Re: [RFC6 PATCH v6 00/21] ILP32 for ARM64
On Tue, 17 May 2016, Szabolcs Nagy wrote:
> i think even legacy software should be able to deal with 64bit off_t,
> so we could avoid having two sets of filesystem apis or is 64bit-only
> off_t more work to do in linux/glibc?
wordsize-64 directories generally expect 64-bit interfaces. wordsize-32
directories generally expect that there are two sets of filesystem APIs
which are not aliased (at the userspace level - the versions for the
generic syscall API deal with setting EOVERFLOW in userspace as needed).
The "wordsize" concept is not wonderfully well-defined and could do with
being split up into multiple better-defined concepts, but that's obvious
something pretty tricky to get right, involving a very careful analysis of
the existing code.
--
Joseph S. Myers
joseph@...esourcery.com
Powered by blists - more mailing lists