From mboxrd@z Thu Jan  1 00:00:00 1970
From: catalin.marinas@arm.com (Catalin Marinas)
Date: Fri, 13 May 2016 12:03:03 +0100
Subject: [RFC6 PATCH v6 00/21] ILP32 for ARM64
In-Reply-To: <20160513105115.GA16672@yury-N73SV>
References: <1459894127-17698-1-git-send-email-ynorov@caviumnetworks.com>
 <20160512002000.GA30997@yury-N73SV>
 <20160512133533.GF11226@e104818-lin.cambridge.arm.com>
 <20160512134431.GB30205@yury-N73SV>
 <20160512140734.GG11226@e104818-lin.cambridge.arm.com>
 <20160512142457.GC30205@yury-N73SV>
 <20160512152805.GJ11226@e104818-lin.cambridge.arm.com>
 <57358C2B.1010106@huawei.com>
 <20160513092802.GA13400@localhost.localdomain>
 <20160513105115.GA16672@yury-N73SV>
Message-ID: <20160513110302.GB27232@e104818-lin.cambridge.arm.com>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

On Fri, May 13, 2016 at 01:51:15PM +0300, Yury Norov wrote:
> On Fri, May 13, 2016 at 09:28:03AM +0000, Catalin Marinas wrote:
> > The discussion is mainly around whether USER_DS for 32-bit compat apps
> > should be the same as USER_DS for native 32-bit apps. Even for native
> > 32-bit kernels, we don't use STACK_TOP as addr_limit. A read/write from
> > 0xffffffff would fail in both cases anyway. I think the LTP test doesn't
> > even try to access such memory but only to probe the range validity (I
> > haven't managed to build the latest LTP yet).
> 
> This fix lets me build it (on top of 7b3ef3b0b)
> Of course, it's not 'official'. :)
[...]
> --- a/testcases/kernel/syscalls/fstatat/fstatat01.c
> +++ b/testcases/kernel/syscalls/fstatat/fstatat01.c
> @@ -59,6 +59,7 @@ static const char *filenames[TEST_CASES];
>  static const int expected_errno[] = { 0, 0, ENOTDIR, EBADF, EINVAL, 0 };
>  static const int flags[] = { 0, 0, 0, 0, 9999, 0 };
>  
> +#define HAVE_FSTATAT
>  #if !defined(HAVE_FSTATAT)
>  #if (__NR_fstatat64 > 0)
>  int fstatat(int dirfd, const char *filename, struct stat64 *statbuf, int flags)
> diff --git a/testcases/kernel/syscalls/preadv/preadv.h b/testcases/kernel/syscalls/preadv/preadv.h
> index f3ac30d..b001389 100644
> --- a/testcases/kernel/syscalls/preadv/preadv.h
> +++ b/testcases/kernel/syscalls/preadv/preadv.h
> @@ -21,6 +21,7 @@
>  #include "config.h"
>  #include "linux_syscall_numbers.h"
>  
> +#define HAVE_PREADV
>  #if !defined(HAVE_PREADV)
>  int preadv(int fd, const struct iovec *iov, int iovcnt, off_t offset)
>  {
> diff --git a/testcases/kernel/syscalls/pwritev/pwritev.h b/testcases/kernel/syscalls/pwritev/pwritev.h
> index ae9d999..2a4d188 100644
> --- a/testcases/kernel/syscalls/pwritev/pwritev.h
> +++ b/testcases/kernel/syscalls/pwritev/pwritev.h
> @@ -21,6 +21,7 @@
>  #include "config.h"
>  #include "linux_syscall_numbers.h"
>  
> +#define HAVE_PWRITEV
>  #if !defined(HAVE_PWRITEV)
>  int pwritev(int fd, const struct iovec *iov, int iovcnt, off_t offset)
>  {

I didn't need any of these defines as they are automatically generated
in include/config.h for AArch32. However, I need to run autoreconf to
generate the configure script prior to building (archlinuxarm
filesystem)

-- 
Catalin