[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-Id: <20180222.141034.147310516964522686.davem@davemloft.net>
Date: Thu, 22 Feb 2018 14:10:34 -0500 (EST)
From: David Miller <davem@...emloft.net>
To: ubraun@...ux.vnet.ibm.com
Cc: netdev@...r.kernel.org, linux-s390@...r.kernel.org,
jwi@...ux.vnet.ibm.com, schwidefsky@...ibm.com,
heiko.carstens@...ibm.com, raspl@...ux.vnet.ibm.com
Subject: Re: [PATCH net-next 2/5] net/smc: fix structure size
From: Ursula Braun <ubraun@...ux.vnet.ibm.com>
Date: Wed, 21 Feb 2018 12:32:32 +0100
> diff --git a/net/smc/smc_cdc.h b/net/smc/smc_cdc.h
> index ab240b37ad11..d2012fd22100 100644
> --- a/net/smc/smc_cdc.h
> +++ b/net/smc/smc_cdc.h
> @@ -48,7 +48,7 @@ struct smc_cdc_msg {
> struct smc_cdc_producer_flags prod_flags;
> struct smc_cdc_conn_state_flags conn_state_flags;
> u8 reserved[18];
> -} __aligned(8);
> +} __packed; /* format defined in RFC7609 */
Hold on, __packed should only be used as the absolute last possible
option to fix structure layout problems.
Also, a sub-structure of smc_cdc_msg, union smc_cdc_cursor, is still
marked with __aligned(8). That makes no sense at all.
Please fix this without using __packed, as __packed has a severe
detrimental effect on code generation for accessing such structure
on several cpu architectures.
Also, if this these are legitimate bug fixes you should target those
at 'net' not 'net-next'.
Thank you.
Powered by blists - more mailing lists