[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <6dcd7c69f95f4fed8cc3d0add1f7d6e9@AcuMS.aculab.com>
Date: Sat, 10 Jun 2023 16:58:54 +0000
From: David Laight <David.Laight@...LAB.COM>
To: David Laight <David.Laight@...LAB.COM>,
'Zhangjin Wu' <falcon@...ylab.org>, "w@....eu" <w@....eu>
CC: "arnd@...db.de" <arnd@...db.de>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"linux-kselftest@...r.kernel.org" <linux-kselftest@...r.kernel.org>,
"linux-riscv@...ts.infradead.org" <linux-riscv@...ts.infradead.org>,
"thomas@...ch.de" <thomas@...ch.de>
Subject: RE: [PATCH v2 4/4] tools/nolibc: sys.h: apply __syscall() helper
From: David Laight
> Sent: 10 June 2023 17:34
>
> From: Zhangjin Wu
> > Sent: 07 June 2023 06:39
> ...
> > As a summary, will use 'sysret()' and something like:
> >
> > static __attribute__((unused))
> > int chdir(const char *path)
> > {
> > return sysret(chdir(path));
> > }
> >
> > to renew the syscall helper patchset, Thanks you very much.
>
> While I'm all for using 'cpp-magic' to abstract and (hopefully)
> simplify things. Token-pasting the sys_ here doesn't seem to gain
> anything.
> Anyone grepping the code for 'sys_chdir' is also going to
> wonder where it is used.
I think I've spotted a later version of the patch
that doesn't paste the sys_
> There might be scope for something like:
> #define syscall_wrapper(func, type) \
> static __attribute__((unused)) \
> int func(type *arg) \
> { \
> return sysret(sys_#func(arg)); \
> }
> and then:
> syscall_wrapper(chdir, const char *)
> would expand to the code above.
>
> I think you'd need separate defines for each number of arguments.
It is also worth pointing out that once the 'static unused'
functions have been defined, all the #define 'goop' used to
define them can be summarily #undef'ed.
That (mostly) solves the namespace problem.
David
-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)
Powered by blists - more mailing lists