All of lore.kernel.org
 help / color / mirror / Atom feed
From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Tim Deegan <tim@xen.org>, xen-devel@lists.xen.org
Cc: Keir Fraser <keir@xen.org>, Jan Beulich <jbeulich@suse.com>
Subject: Re: [PATCH] x86_emulate: split the {reg, mem} union in struct operand.
Date: Thu, 26 Mar 2015 12:25:55 +0000	[thread overview]
Message-ID: <5513FAD3.9050400@citrix.com> (raw)
In-Reply-To: <1427372174-980-1-git-send-email-tim@xen.org>

On 26/03/15 12:16, Tim Deegan wrote:
> In the hopes of making any future errors along the lines of XSA-123
> into clean crashes instead of memory corruption bugs.
>
> Signed-off-by: Tim Deegan <tim@xen.org>
> CC: Keir Fraser <keir@xen.org>
> CC: Jan Beulich <jbeulich@suse.com>
> CC: Andrew Cooper <andrew.cooper3@citrix.com>

I am very much in favour of this change.

Does REG_POISON work with a 32bit compile of the test harness?

Independent of this, it might be easier to spot in crashes if the upper 
word of the address was also 8086

~Andrew

> ---
>   xen/arch/x86/x86_emulate/x86_emulate.c | 8 +++++---
>   1 file changed, 5 insertions(+), 3 deletions(-)
>
> diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c
> index c082c9e..37d1fea 100644
> --- a/xen/arch/x86/x86_emulate/x86_emulate.c
> +++ b/xen/arch/x86/x86_emulate/x86_emulate.c
> @@ -319,7 +319,7 @@ struct operand {
>       /* Original operand value. */
>       unsigned long orig_val;
>   
> -    union {
> +    struct {
>           /* OP_REG: Pointer to register field. */
>           unsigned long *reg;
>           /* OP_MEM: Segment and offset. */
> @@ -329,6 +329,7 @@ struct operand {
>           } mem;
>       };
>   };
> +#define REG_POISON ((unsigned long *) 0x8000000000008086UL) /* non-canonical */
>   
>   typedef union {
>       uint64_t mmx;
> @@ -1447,14 +1448,15 @@ x86_emulate(
>       unsigned int op_bytes, def_op_bytes, ad_bytes, def_ad_bytes;
>       bool_t lock_prefix = 0;
>       int override_seg = -1, rc = X86EMUL_OKAY;
> -    struct operand src, dst;
> +    struct operand src = { .reg = REG_POISON };
> +    struct operand dst = { .reg = REG_POISON };
>       enum x86_swint_type swint_type;
>       DECLARE_ALIGNED(mmval_t, mmval);
>       /*
>        * Data operand effective address (usually computed from ModRM).
>        * Default is a memory operand relative to segment DS.
>        */
> -    struct operand ea = { .type = OP_MEM };
> +    struct operand ea = { .type = OP_MEM, .reg = REG_POISON };
>       ea.mem.seg = x86_seg_ds; /* gcc may reject anon union initializer */
>   
>       ctxt->retire.byte = 0;

  reply	other threads:[~2015-03-26 12:25 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-03-26 12:16 [PATCH] x86_emulate: split the {reg, mem} union in struct operand Tim Deegan
2015-03-26 12:25 ` Andrew Cooper [this message]
2015-03-26 12:34   ` Tim Deegan
2015-03-26 12:40 ` [PATCH v2] " Tim Deegan
2015-03-26 12:42   ` Andrew Cooper
2015-03-27 14:31   ` Jan Beulich
2015-03-27 14:44     ` [PATCH v3] " Tim Deegan

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=5513FAD3.9050400@citrix.com \
    --to=andrew.cooper3@citrix.com \
    --cc=jbeulich@suse.com \
    --cc=keir@xen.org \
    --cc=tim@xen.org \
    --cc=xen-devel@lists.xen.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.