[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAEdQ38FNiY78zXMtqCUCjiGYBfjHXnDTQia3a62hJ_6UG3O+gw@mail.gmail.com>
Date: Fri, 2 Nov 2018 12:14:50 -0700
From: Matt Turner <mattst88@...il.com>
To: jrdr.linux@...il.com
Cc: Michael Cree <mcree@...on.net.nz>, brajeswar.linux@...il.com,
Richard Henderson <rth@...ddle.net>,
Ivan Kokshaysky <ink@...assic.park.msu.ru>,
Andrea Parri <parri.andrea@...il.com>,
Ingo Molnar <mingo@...nel.org>,
"Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>,
linux-alpha <linux-alpha@...r.kernel.org>,
LKML <linux-kernel@...r.kernel.org>, sabyasachi.linux@...il.com
Subject: Re: [PATCH] include/asm/cmpxchg.h: Remove duplicate header
On Fri, Nov 2, 2018 at 12:09 PM Souptick Joarder <jrdr.linux@...il.com> wrote:
>
> On Sat, Nov 3, 2018 at 12:31 AM Matt Turner <mattst88@...il.com> wrote:
> >
> > On Fri, Nov 2, 2018 at 11:55 AM Souptick Joarder <jrdr.linux@...il.com> wrote:
> > > We run the static analyser "make includecheck" which list out files where
> > > duplicate headers can be removed and based on that we thought to remove
> > > from this file. Didn't understood about the existence of second include ??
> >
> >
> > #define ____xchg(type, args...) __xchg ## type ## _local(args)
> > #define ____cmpxchg(type, args...) __cmpxchg ## type ## _local(args)
> > #include <asm/xchg.h>
> > [snip]
> > #undef ____xchg
> > #undef ____cmpxchg
> > #define ____xchg(type, args...) __xchg ##type(args)
> > #define ____cmpxchg(type, args...) __cmpxchg ##type(args)
> > #include <asm/xchg.h>
> >
> > asm/xchg.h has a comment at the top that says
> >
> > /*
> > * xchg/xchg_local and cmpxchg/cmpxchg_local share the same code
> > * except that local version do not have the expensive memory barrier.
> > * So this file is included twice from asm/cmpxchg.h.
> > */
>
> Thanks Matt. Sorry for the noise.
> Is there any way to exclude it from static analyser that someone else will
> not do the same mistake in future ?
Since this is not an uncommon pattern in C, I think any static
analysis tool that attempts to find duplicate includes should attempt
to recognize such a pattern.
That, or humans should review the output of their static analysis
tools. Or you could try to compile the patches produced. I think any
of those would have caught the problem with the patch.
Powered by blists - more mailing lists