Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Fri, 18 Feb 2022 17:15:16 -0600
From:   "Gustavo A. R. Silva" <gustavoars@kernel.org>
To:     Randy Dunlap <rdunlap@infradead.org>
Cc:     "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: fallthrough question
Message-ID: <20220218231516.GA1091555@embeddedor>
References: <858a2b13-64e1-99bc-e2f7-5b2477bb1126@infradead.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <858a2b13-64e1-99bc-e2f7-5b2477bb1126@infradead.org>
Precedence: bulk

Hi there,

On Fri, Feb 18, 2022 at 01:57:58PM -0800, Randy Dunlap wrote:
> Hi,
> 
> I expected this to produce a fallthrough warning, but it doesn't
> (with gcc 11.1.0):
> 
> from sound/oss/dmasound/dmasound_core.c:#1481, when falling from case 1
> into case 0: (in arch/m68k/ selected builds only)
> 
> 	case 1:
> 		if ((size = ints[2]) < 256) /* check for small buffer specs */
> 			size <<= 10 ;
>                 if (size < MIN_BUFSIZE || size > MAX_BUFSIZE)
>                         printk("dmasound_setup: invalid write buffer size, using default = %d\n", writeBufSize);
>                 else
>                         writeBufSize = size;
> 	case 0:
> 		break;
> 	default:
> 
> Can you tell me what is going on here?

As you can see the warning is suppressed when a case label falls through to
a case that merely breaks... or returns, or continues (continue statement)
or that goes to (goto statement) some other place.

However, Clang disagrees with this. See below for more:

https://github.com/ClangBuiltLinux/linux/issues/636
https://godbolt.org/z/xgkvIh

--
Gustavo