qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Richard Henderson <richard.henderson@linaro.org>
To: David Hildenbrand <david@redhat.com>, qemu-devel@nongnu.org
Cc: Florian Weimer <fweimer@redhat.com>,
	Thomas Huth <thuth@redhat.com>, Cornelia Huck <cohuck@redhat.com>,
	Stefano Brivio <sbrivio@redhat.com>,
	qemu-s390x@nongnu.org, Richard Henderson <rth@twiddle.net>
Subject: Re: [Qemu-devel] [PATCH v2 04/28] s390x/tcg: MVCL: Process max 2k bytes at a time
Date: Wed, 11 Sep 2019 10:52:36 -0400	[thread overview]
Message-ID: <7788b1b6-fbf0-3370-5563-1da4f7124999@linaro.org> (raw)
In-Reply-To: <20190906075750.14791-5-david@redhat.com>

On 9/6/19 3:57 AM, David Hildenbrand wrote:
> Process max 2k bytes at a time, writing back registers between the
> accesses. The instruction is interruptible.
>     "For operands longer than 2K bytes, access exceptions are not
>     recognized for locations more than 2K bytes beyond the current location
>     being processed."
> 
> MVCL handling is quite different than MVCLE/MVCLU handling, so split up
> the handlers.
> 
> We'll deal with fast_memmove() and fast_memset() not probing
> reads/writes properly later. Also, we'll defer interrupt handling, as
> that will require more thought, add a TODO for that.
> 
> Signed-off-by: David Hildenbrand <david@redhat.com>
> ---
>  target/s390x/mem_helper.c | 44 +++++++++++++++++++++++++++++++++------
>  1 file changed, 38 insertions(+), 6 deletions(-)
> 
> diff --git a/target/s390x/mem_helper.c b/target/s390x/mem_helper.c
> index 2361ed6d54..2e22c183bd 100644
> --- a/target/s390x/mem_helper.c
> +++ b/target/s390x/mem_helper.c
> @@ -799,19 +799,51 @@ uint32_t HELPER(mvcl)(CPUS390XState *env, uint32_t r1, uint32_t r2)
>      uint64_t srclen = env->regs[r2 + 1] & 0xffffff;
>      uint64_t src = get_address(env, r2);
>      uint8_t pad = env->regs[r2 + 1] >> 24;
> -    uint32_t cc;
> +    uint32_t cc, cur_len;
>  
>      if (is_destructive_overlap(env, dest, src, MIN(srclen, destlen))) {
>          cc = 3;
> +    } else if (srclen == destlen) {
> +        cc = 0;
> +    } else if (destlen < srclen) {
> +        cc = 1;
>      } else {
> -        cc = do_mvcl(env, &dest, &destlen, &src, &srclen, pad, 1, ra);
> +        cc = 2;
>      }
>  
> -    env->regs[r1 + 1] = deposit64(env->regs[r1 + 1], 0, 24, destlen);
> -    env->regs[r2 + 1] = deposit64(env->regs[r2 + 1], 0, 24, srclen);
> -    set_address_zero(env, r1, dest);
> -    set_address_zero(env, r2, src);
> +    /* We might have to zero-out some bits even if there was no action. */
> +    if (unlikely(!destlen || cc == 3)) {
> +        set_address_zero(env, r2, src);
> +        set_address_zero(env, r1, dest);
> +        return cc;
> +    } else if (!srclen) {
> +        set_address_zero(env, r2, src);
> +    }
>  
> +    /*
> +     * Only perform one type of type of operation (move/pad) in one step.
> +     * Process up to 2k bytes.
> +     */
> +    while (destlen) {
> +        cur_len = MIN(destlen, 2048);

The language in the PoP is horribly written, and thus confusing.  I can't
believe it really means what it appears to say, about access exceptions not
being recognized.

The code within Hercules breaks the action at every 2k address boundary -- for
both src and dest.  That's the only way that actually makes sense to me, as
otherwise we end up allowing userspace to read/write into a page without
permission.  Which is a security hole.


r~


  reply	other threads:[~2019-09-11 15:04 UTC|newest]

Thread overview: 68+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-09-06  7:57 [Qemu-devel] [PATCH v2 00/28] s390x/tcg: mem_helper: Fault-safe handling David Hildenbrand
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 01/28] s390x/tcg: Reset exception_index to -1 instead of 0 David Hildenbrand
2019-09-11 14:38   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 02/28] s390x/tcg: MVCL: Zero out unused bits of address David Hildenbrand
2019-09-11 14:40   ` Richard Henderson
2019-09-11 16:10     ` David Hildenbrand
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 03/28] s390x/tcg: MVCL: Detect destructive overlaps David Hildenbrand
2019-09-11 14:42   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 04/28] s390x/tcg: MVCL: Process max 2k bytes at a time David Hildenbrand
2019-09-11 14:52   ` Richard Henderson [this message]
2019-09-11 15:07     ` Richard Henderson
2019-09-11 16:12       ` David Hildenbrand
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 05/28] s390x/tcg: MVC: Increment the length once David Hildenbrand
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 06/28] s390x/tcg: MVC: Use is_destructive_overlap() David Hildenbrand
2019-09-11 14:54   ` Richard Henderson
2019-09-11 16:13     ` David Hildenbrand
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 07/28] s390x/tcg: MVPG: Check for specification exceptions David Hildenbrand
2019-09-11 14:57   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 08/28] s390x/tcg: MVPG: Properly wrap the addresses David Hildenbrand
2019-09-11 14:58   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 09/28] s390x/tcg: MVCLU/MVCLE: Process max 4k bytes at a time David Hildenbrand
2019-09-11 15:05   ` Richard Henderson
2019-09-11 16:14     ` David Hildenbrand
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 10/28] s390x/tcg: MVCS/MVCP: Check for special operation exceptions David Hildenbrand
2019-09-11 15:08   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 11/28] s390x/tcg: MVCS/MVCP: Properly wrap the length David Hildenbrand
2019-09-11 15:11   ` Richard Henderson
2019-09-11 16:15     ` David Hildenbrand
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 12/28] s390x/tcg: MVST: Check for specification exceptions David Hildenbrand
2019-09-11 15:14   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 13/28] s390x/tcg: MVST: Fix storing back the addresses to registers David Hildenbrand
2019-09-11 15:18   ` Richard Henderson
2019-09-11 16:15     ` David Hildenbrand
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 14/28] s390x/tcg: Always use MMU_USER_IDX for CONFIG_USER_ONLY David Hildenbrand
2019-09-11 15:19   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 15/28] s390x/tcg: Fault-safe memset David Hildenbrand
2019-09-11 15:29   ` Richard Henderson
2019-09-11 16:18     ` David Hildenbrand
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 16/28] s390x/tcg: Fault-safe memmove David Hildenbrand
2019-09-11 21:11   ` Richard Henderson
2019-09-11 22:03   ` Richard Henderson
2019-09-13 12:37     ` David Hildenbrand
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 17/28] s390x/tcg: MVCS/MVCP: Use access_memmove_idx() David Hildenbrand
2019-09-11 21:13   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 18/28] s390x/tcg: MVC: Fault-safe handling on destructive overlaps David Hildenbrand
2019-09-11 21:20   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 19/28] s390x/tcg: MVCLU: Fault-safe handling David Hildenbrand
2019-09-11 21:24   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 20/28] s390x/tcg: OC: " David Hildenbrand
2019-09-11 21:26   ` Richard Henderson
2019-09-16 12:01     ` David Hildenbrand
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 21/28] s390x/tcg: XC: " David Hildenbrand
2019-09-11 21:29   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 22/28] s390x/tcg: NC: " David Hildenbrand
2019-09-11 21:32   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 23/28] s390x/tcg: MVCIN: " David Hildenbrand
2019-09-11 21:35   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 24/28] s390x/tcg: MVN: " David Hildenbrand
2019-09-11 21:37   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 25/28] s390x/tcg: MVZ: " David Hildenbrand
2019-09-11 21:38   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 26/28] s390x/tcg: MVST: " David Hildenbrand
2019-09-11 21:52   ` Richard Henderson
2019-09-16 10:39     ` David Hildenbrand
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 27/28] s390x/tcg: MVO: " David Hildenbrand
2019-09-11 22:09   ` Richard Henderson
2019-09-06  7:57 ` [Qemu-devel] [PATCH v2 28/28] tests/tcg: target/s390x: Test MVO David Hildenbrand
2019-09-11 11:11 ` [Qemu-devel] [qemu-s390x] [PATCH v2 00/28] s390x/tcg: mem_helper: Fault-safe handling David Hildenbrand

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=7788b1b6-fbf0-3370-5563-1da4f7124999@linaro.org \
    --to=richard.henderson@linaro.org \
    --cc=cohuck@redhat.com \
    --cc=david@redhat.com \
    --cc=fweimer@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=qemu-s390x@nongnu.org \
    --cc=rth@twiddle.net \
    --cc=sbrivio@redhat.com \
    --cc=thuth@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).