From mboxrd@z Thu Jan  1 00:00:00 1970
From: Petr Vorel <pvorel@suse.cz>
Date: Wed, 26 May 2021 12:32:48 +0200
Subject: [LTP] [PATCH 0/2 v2] syscalls/sendfile: Convert sendfile{08,
 09} to the new API
In-Reply-To: <YK4Q1iJz0/i4ZSxc@yuki>
References: <20210520092449.162043-1-xieziyao@huawei.com>
 <YK0XBktEIyGkhAsT@pevik> <YK4Q1iJz0/i4ZSxc@yuki>
Message-ID: <YK4j0Pw4ECqjFWTj@pevik>
List-Id: <ltp.lists.linux.it>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: ltp@lists.linux.it

Hi Cyril,

> Hi!
> > By accident, I changed
> > off_t before_pos, after_pos;
> > to:
> > OFF_T before_pos, after_pos;

> > @Cyril: should it be reverted?

> There is a define in the Makefile that defines OFF_T to off64_t for
> _FILE_OFFSET_BITS=64 but that looks like a workaround for buggy libc.

> The _FILE_OFFSET_BITS=64 should, when code is compiled on 32bit
> platform, change the 32bit off_t to a 64bit off_t.
Thanks for info. I noticed Makefile myself, but didn't know the background.

> I.e. when you compile following:

> int main(void)
> {
>         printf("%zu\n", sizeof(off_t));
>         return 0;
> }

> on a 32bit (or with -m32) it should print 4
> on a 32bit with -D_FILE_OFFSET_BITS=64 it should print 8
> on a 64bit it should print 8 regardless
Yep (verified myself).

> As far as I can tell we should remove the OFF_T handling from the
> Makefile here instead.
OK, I'll send a patch which removes off64_t from Makefile and use
always off_t.

Kind regards,
Petr