linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Arnd Bergmann <arnd@kernel.org>
To: Don.Brace@microchip.com
Cc: Sergei Trofimovich <slyfox@gentoo.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	linux-ia64@vger.kernel.org, storagedev@microchip.com,
	linux-scsi <linux-scsi@vger.kernel.org>,
	jszczype@redhat.com, Scott.Benesh@microchip.com,
	Scott.Teel@microchip.com, thenzl@redhat.com,
	"Martin K. Petersen" <martin.petersen@oracle.com>,
	John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Subject: Re: [PATCH] hpsa: fix boot on ia64 (atomic_t alignment)
Date: Tue, 16 Mar 2021 19:28:22 +0100	[thread overview]
Message-ID: <CAK8P3a3JYmhbN3TXB2cWGfXGKgsUa9Hg=ZvWckTaL_31OMgbtQ@mail.gmail.com> (raw)
In-Reply-To: <SN6PR11MB2848561E3D85A8F55EB86977E16B9@SN6PR11MB2848.namprd11.prod.outlook.com>

On Tue, Mar 16, 2021 at 6:12 PM <Don.Brace@microchip.com> wrote:

>  drivers/scsi/hpsa_cmd.h | 14 +++++++++++++-
>  1 file changed, 13 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/scsi/hpsa_cmd.h b/drivers/scsi/hpsa_cmd.h index d126bb877250..617bdae9a7de 100644
> --- a/drivers/scsi/hpsa_cmd.h
> +++ b/drivers/scsi/hpsa_cmd.h
> @@ -20,6 +20,9 @@
>  #ifndef HPSA_CMD_H
>  #define HPSA_CMD_H
>
> +#include <linux/build_bug.h> /* static_assert */ #include
> +<linux/stddef.h> /* offsetof */
> +
>  /* general boundary defintions */
>  #define SENSEINFOBYTES          32 /* may vary between hbas */
>  #define SG_ENTRIES_IN_CMD      32 /* Max SG entries excluding chain blocks */
> @@ -448,11 +451,20 @@ struct CommandList {
>          */
>         struct hpsa_scsi_dev_t *phys_disk;
>
> -       bool retry_pending;
> +       int retry_pending;
>         struct hpsa_scsi_dev_t *device;
>         atomic_t refcount; /* Must be last to avoid memset in hpsa_cmd_init() */  } __aligned(COMMANDLIST_ALIGNMENT);
>
> +/*
> + * Make sure our embedded atomic variable is aligned. Otherwise we
> +break atomic
> + * operations on architectures that don't support unaligned atomics like IA64.
> + *
> + * Ideally this header should be cleaned up to only mark individual
> +structs as
> + * packed.
> + */
> +static_assert(offsetof(struct CommandList, refcount) %
> +__alignof__(atomic_t) == 0);
> +

Actually that still feels wrong: the annotation of the struct is to pack
every member, which causes the access to be done in byte units
on architectures that do not have hardware unaligned load/store
instructions, at least for things like atomic_read() that does not
go through a cmpxchg() or ll/sc cycle.

This change may fix itanium, but it's still not correct. Other
architectures would have already been broken before the recent
change, but that's not a reason against fixing them now.

I'd recommend marking the entire structure as having default
alignment, by adding the appropriate pragmas before and after it.

         Arnd

  reply	other threads:[~2021-03-16 18:29 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <20210222230519.73f3e239@sf>
2021-02-22 23:34 ` 5.11 regression: "ia64: add support for TIF_NOTIFY_SIGNAL" breaks ia64 boot Jens Axboe
2021-02-22 23:55   ` John Paul Adrian Glaubitz
     [not found]     ` <20210223083507.43b5a6dd@sf>
2021-02-23  9:13       ` John Paul Adrian Glaubitz
2021-02-23 12:36         ` John Paul Adrian Glaubitz
     [not found]           ` <20210223192743.0198d4a9@sf>
     [not found]             ` <20210302222630.5056f243@sf>
2021-03-02 22:31               ` John Paul Adrian Glaubitz
2021-03-03  0:22                 ` [bisected] 5.12-rc1 hpsa regression: "scsi: hpsa: Correct dev cmds outstanding for retried cmds" breaks hpsa P600 Sergei Trofimovich
2021-03-03  8:55                   ` Sergei Trofimovich
2021-03-03 17:33                     ` Don.Brace
     [not found]                       ` <20210303220401.501449e5@sf>
2021-03-04 17:00                         ` Don.Brace
2021-03-05 13:26                           ` Tomas Henzl
2021-03-12 22:27                             ` [PATCH] hpsa: fix boot on ia64 (atomic_t alignment) Sergei Trofimovich
2021-03-16 16:30                               ` Don.Brace
2021-03-16 18:28                                 ` Arnd Bergmann [this message]
2021-03-17  2:25                                   ` Martin K. Petersen
2021-03-17 13:19                                     ` David Laight
2021-03-17 19:06                                       ` Don.Brace
2021-03-17 17:28                               ` John Paul Adrian Glaubitz
2021-03-27 10:24                                 ` Sergei Trofimovich
2021-03-24  7:08                               ` John Paul Adrian Glaubitz
2021-03-24 18:37                               ` Don.Brace
2021-03-29 11:25                                 ` John Paul Adrian Glaubitz
2021-03-29 14:22                                   ` Arnd Bergmann
2021-03-30  3:02                                     ` Martin K. Petersen
2021-03-30  7:19                                       ` [PATCH v2 1/3] hpsa: use __packed on individual structs, not header-wide Sergei Trofimovich
2021-03-30  7:19                                         ` [PATCH v2 2/3] hpsa: fix boot on ia64 (atomic_t alignment) Sergei Trofimovich
2021-03-30  7:19                                         ` [PATCH v2 3/3] hpsa: add an assert to prevent from __packed reintroduction Sergei Trofimovich
2021-03-30  7:34                                           ` Arnd Bergmann
2021-04-02 14:40                                             ` Elliott, Robert (Servers)
2021-04-03 14:51                                               ` Sergei Trofimovich
2021-03-30  7:30                                         ` [PATCH v2 1/3] hpsa: use __packed on individual structs, not header-wide Arnd Bergmann
2021-03-30  7:43                                           ` Arnd Bergmann
2021-04-02  3:54                                         ` Martin K. Petersen
2021-04-15 18:41                                           ` Don.Brace
2021-03-05  9:22                       ` [bisected] 5.12-rc1 hpsa regression: "scsi: hpsa: Correct dev cmds outstanding for retried cmds" breaks hpsa P600 Geert Uytterhoeven
2021-03-05 13:31                         ` Arnd Bergmann
2021-03-05 20:45                           ` Don.Brace
2021-03-03 15:42                   ` Don.Brace
2021-03-17 17:42             ` 5.11 regression: "ia64: add support for TIF_NOTIFY_SIGNAL" breaks ia64 boot John Paul Adrian Glaubitz
2021-03-17 17:53               ` John Paul Adrian Glaubitz
     [not found]   ` <20210222235359.75d1a912@sf>
2021-02-23  0:34     ` Jens Axboe
2021-02-23  0:41       ` Jens Axboe
2021-02-23  0:43         ` Jens Axboe
     [not found]           ` <20210223080830.23bccdbf@sf>
2021-03-02 22:07             ` Sergei Trofimovich
2021-03-02 22:31               ` Jens Axboe
     [not found]                 ` <20210302232716.353ed49b@sf>
2021-03-03  0:34                   ` Jens Axboe
2021-03-03  3:51                     ` Jens Axboe

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAK8P3a3JYmhbN3TXB2cWGfXGKgsUa9Hg=ZvWckTaL_31OMgbtQ@mail.gmail.com' \
    --to=arnd@kernel.org \
    --cc=Don.Brace@microchip.com \
    --cc=Scott.Benesh@microchip.com \
    --cc=Scott.Teel@microchip.com \
    --cc=glaubitz@physik.fu-berlin.de \
    --cc=jszczype@redhat.com \
    --cc=linux-ia64@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-scsi@vger.kernel.org \
    --cc=martin.petersen@oracle.com \
    --cc=slyfox@gentoo.org \
    --cc=storagedev@microchip.com \
    --cc=thenzl@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).