[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <e1259039a99d4f23958c4380e5c0b11f@AcuMS.aculab.com>
Date: Sun, 8 Sep 2024 20:08:23 +0000
From: David Laight <David.Laight@...LAB.COM>
To: 'Andy Shevchenko' <andy.shevchenko@...il.com>, zhangjiao2
<zhangjiao2@...s.chinamobile.com>
CC: "wbg@...nel.org" <wbg@...nel.org>, "linux-iio@...r.kernel.org"
<linux-iio@...r.kernel.org>, "linux-kernel@...r.kernel.org"
<linux-kernel@...r.kernel.org>
Subject: RE: [PATCH] tools/counter: close fd when exit
From: Andy Shevchenko
> Sent: 04 September 2024 00:20
>
> Tue, Sep 03, 2024 at 04:01:21PM +0800, zhangjiao2 kirjoitti:
> > From: zhang jiao <zhangjiao2@...s.chinamobile.com>
> >
> > close fd when exit the program
>
> Please, mind English grammar and punctuation.
> Also this doesn't state what the problem is.
After all, when the program exits the kernel closes all the files.
> > if (ret == -1) {
> > fprintf(stderr, "Error adding watches[%d]: %s\n", i,
> > strerror(errno));
> > + close(fd);
>
> Since fd is not used in the messaging it's better to close it before printing
> anything. Ditto for other cases.
Unless close() decides to change errno...
>
> ...
>
> > if (ret != sizeof(event_data)) {
> > fprintf(stderr, "Failed to read event data\n");
> > + close(fd);
>
> > return -EIO;
>
> Side note: This will return garbage to the userspace. Should be
>
> return EIO;
If that is the return value from main() it is equally invalid.
Valid return values are 0..255 but 128..255 are likely to be
interpreted as 'killed by a signal'.
So really you only get 0..127.
Mostly used as 0 => success and 1 => failure.
David
-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)
Powered by blists - more mailing lists