This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [RFC6 PATCH v6 00/21] ILP32 for ARM64
- From: Arnd Bergmann <arnd at arndb dot de>
- To: Andreas Schwab <schwab at suse dot de>
- Cc: Joseph Myers <joseph at codesourcery dot com>, Szabolcs Nagy <szabolcs dot nagy at arm dot com>, Yury Norov <ynorov at caviumnetworks dot com>, catalin dot marinas at arm dot com, linux-arm-kernel at lists dot infradead dot org, linux-kernel at vger dot kernel dot org, nd at arm dot com, linux-arch at vger dot kernel dot org, linux-s390 at vger dot kernel dot org, pinskia at gmail dot com, Prasun dot Kapoor at caviumnetworks dot com, broonie at kernel dot org, linux-doc at vger dot kernel dot org, heiko dot carstens at de dot ibm dot com, agraf at suse dot de, klimov dot linux at gmail dot com, bamvor dot zhangjian at huawei dot com, schwidefsky at de dot ibm dot com, Nathan_Lynch at mentor dot com, christoph dot muellner at theobroma-systems dot com, GNU C Library <libc-alpha at sourceware dot org>
- Date: Wed, 18 May 2016 00:45:51 +0200
- Subject: Re: [RFC6 PATCH v6 00/21] ILP32 for ARM64
- Authentication-results: sourceware.org; auth=none
- References: <1459894127-17698-1-git-send-email-ynorov at caviumnetworks dot com> <alpine dot DEB dot 2 dot 20 dot 1605171542340 dot 24255 at digraph dot polyomino dot org dot uk> <mvm1t50ad0j dot fsf at hawking dot suse dot de>
On Tuesday 17 May 2016 18:02:36 Andreas Schwab wrote:
> Joseph Myers <joseph@codesourcery.com> writes:
>
> > On Tue, 17 May 2016, Arnd Bergmann wrote:
> >
> >> I think it has become easier to override now and we just need to
> >> update sysdeps/unix/sysv/linux/generic/bits/typesizes.h to set
> >>
> >> #define __INO64_T_TYPE __UQUAD_TYPE
> >> #define __OFF64_T_TYPE __UQUAD_TYPE
> >> #define __OFF_T_MATCHES_OFF64_T 1
> >> #define __INO_T_MATCHES_INO64_T 1
> >>
> >> for new architectures (obviously not the ones that already use the
> >> 32-bit types). I haven't tries this, so there may be other things
> >> that are required.
> >
> > I think more than that would be needed to get struct stat to match and get
> > things aliased for that (which is presumably desirable).
>
> Looking at sysdeps/unix/sysv/linux/generic/bits/stat.h, there is at
> least blkcnt_t that differs between stat and stat64. And you probably
> want to alias statfs and statfs64 as well, ie. fs{blk,fil}cnt_t.
Makes sense. Indeed that is a bit more work than I hoped, in particular
if we have to audit the uses of __WORDSIZE as well.
Arnd