linux-kernel - Re: [RFC PATCH 35/35] syscalls: do not call sys

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20180312194311.GE16666@light.dominikbrodowski.net>
Date:   Mon, 12 Mar 2018 20:43:11 +0100
From:   Dominik Brodowski <linux@...inikbrodowski.net>
To:     Ingo Molnar <mingo@...nel.org>
Cc:     linux-kernel@...r.kernel.org, luto@...nel.org,
        torvalds@...ux-foundation.org, viro@...iv.linux.org.uk,
        akpm@...ux-foundation.org
Subject: Re: [RFC PATCH 35/35] syscalls: do not call sys_close() within the
 kernel

On Mon, Mar 12, 2018 at 08:37:19AM +0100, Ingo Molnar wrote:
> 
> * Dominik Brodowski <linux@...inikbrodowski.net> wrote:
> 
> > --- a/fs/open.c
> > +++ b/fs/open.c
> > @@ -1200,7 +1200,7 @@ SYSCALL_DEFINE1(close, unsigned int, fd)
> >  
> >  	return retval;
> >  }
> > -EXPORT_SYMBOL(sys_close);
> > +
> >  
> >  /*
> 
> Nit: this introduces a stray newline.
> 
> > --- a/include/linux/syscalls.h
> > +++ b/include/linux/syscalls.h
> > @@ -1042,4 +1042,10 @@ static inline long ksys_lchown(const char __user *filename, uid_t user,
> >  	return do_fchownat(AT_FDCWD, filename, user, group,
> >  			     AT_SYMLINK_NOFOLLOW);
> >  }
> > +
> > +extern int __close_fd(struct files_struct *files, unsigned int fd);
> > +static inline int ksys_close(unsigned int fd)
> > +{
> > +	return __close_fd(current->files, fd);
> > +}
> 
> Would be nice to reuse that stray newline after the __close_fd() prototype for the 
> canonical stylistic separation of declarations from definitions.
> 
> It would also be very nice to add a short comment before ksys_close() that 
> explains how it differs from sys_close(). This should reduce the amount of 
> cargo-cult copying of existing ksys_close()/sys_close() patterns.

Thanks, have done that:

	/*
	 * In contrast to sys_close(), this stub does not check whether the syscall
	 * should or should not be restarted, but returns the raw error codes from
	 * __close_fd().
	 */

Is that ok?

Thanks,
	Dominik