From mboxrd@z Thu Jan  1 00:00:00 1970
From: Steve Muckle <smuckle@google.com>
Date: Fri, 1 Feb 2019 16:19:07 -0800
Subject: [LTP] [PATCH] syscalls/ustat: convert to new lib,
 use direct syscall
In-Reply-To: <20190130104007.GA32120@rei>
References: <20190129023714.191820-1-smuckle@google.com>
 <20190129180617.GD14252@dell5510> <20190130104007.GA32120@rei>
Message-ID: <5468c7b4-0236-2637-9bac-e207dfe97e5b@google.com>
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

On 01/30/2019 02:40 AM, Cyril Hrubis wrote:
> Hi!
>>> +++ b/include/lapi/ustat.h
>>> @@ -0,0 +1,8 @@
>>> +#include <sys/types.h>
>> I'd put here at least
>> //SPDX-License-Identifier: GPL-2.0-or-later
>>> +
>>> +struct ustat {
>>> +	daddr_t f_tfree;
>>> +	ino_t f_tinode;
>>> +	char f_fname[6];
>>> +	char f_fpack[6];
>>> +};
>> Thinking about it twice, wouldn't be better to include <sys/ustat.h> and detect
>> with autotools whether struct ustat is defined and use this only if it's not
>> defined? The same approach we use for struct statx?
> 
> And we do have the configure check for ustat.h in place so this should
> be as easy as doing:
> 
> #ifdef HAVE_SYS_USTAT_H
> # include <sys/ustat.h>
> #else
> struct ustat {
> ...
> }
> #endif
> 
> Also ideally this header should include guards, even if it's unlikely to
> be included twice.
> 
>>> +++ b/testcases/kernel/syscalls/ustat/ustat01.c
>>> @@ -1,21 +1,7 @@
>>> +// SPDX-License-Identifier: GPL-2.0
>> @Cyril: can we use GPL-2.0-or-later? Original was GPL only.
> 
> I would have kept it GPL-2.0 unless it's rewritten from scratch
> completely.

Thanks Petr and Cyril for the review, v2 sent.