Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Message-ID: <3bf931c6ea0cae3e23f3485801986859851b4f04.camel@perches.com>
Subject: Re: [PATCH] lib/lz4: smatch warning in LZ4_decompress_generic()
From:   Joe Perches <joe@perches.com>
To:     Yann Collet <cyan@fb.com>, Vasily Averin <vvs@virtuozzo.com>,
        Gao Xiang <hsiangkao@aol.com>
Cc:     "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        Andrew Morton <akpm@linux-foundation.org>,
        Gao Xiang <xiang@kernel.org>
Date:   Sun, 07 Jun 2020 18:02:44 -0700
In-Reply-To: <D4762145-BBC5-4574-BF68-8C1A3AF41D98@fb.com>
References: <20200606143646.GB10839.ref@hsiangkao-HP-ZHAN-66-Pro-G1>
         <20200606143646.GB10839@hsiangkao-HP-ZHAN-66-Pro-G1>
         <330eccf1-6d4b-07dd-4e55-ffe3a179e4b8@virtuozzo.com>
         <D4762145-BBC5-4574-BF68-8C1A3AF41D98@fb.com>
Content-Type: text/plain; charset="ISO-8859-1"
User-Agent: Evolution 3.36.2-0ubuntu1 
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, 2020-06-08 at 00:40 +0000, Yann Collet wrote:
> Hi Vasily
> 
> 
> If I do understand the discussion, the question is about usage of `&` instead of `&&`,
> and the speculation that it might be an error.
> 
> It's not an error. Unfortunately, explaining the reasoning behind this decision is a bit long.

Likely better to add a comment around the use so that
another patch like this doesn't get submitted again.

Perhaps something like:
---
 lib/lz4/lz4_decompress.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/lib/lz4/lz4_decompress.c b/lib/lz4/lz4_decompress.c
index 0c9d3ad17e0f..5371dab6b481 100644
--- a/lib/lz4/lz4_decompress.c
+++ b/lib/lz4/lz4_decompress.c
@@ -141,6 +141,9 @@ static FORCE_INLINE int LZ4_decompress_generic(
 		 * space in the output for those 18 bytes earlier, upon
 		 * entering the shortcut (in other words, there is a
 		 * combined check for both stages).
+		 *
+		 * The & in the likely() below is intentionally not && so that
+		 * some compilers can produce better parallelized runtime code
 		 */
 		if ((endOnInput ? length != RUN_MASK : length <= 8)
 		   /*