public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: "H.J. Lu" <hjl.tools@gmail.com>
To: Paolo Bonzini <bonzini@gnu.org>
Cc: GCC Patches <gcc-patches@gcc.gnu.org>
Subject: Re: PING: PATCH [4/n]: Prepare x32: Permute the conversion and addition if one operand is a constant
Date: Mon, 11 Jul 2011 15:58:00 -0000	[thread overview]
Message-ID: <CAMe9rOp7=UvknX2z9k5O1Q8sJkXH8Z11=7TEjU0M5_EoAFC11g@mail.gmail.com> (raw)
In-Reply-To: <4E1AD86A.8000800@gnu.org>

On Mon, Jul 11, 2011 at 4:03 AM, Paolo Bonzini <bonzini@gnu.org> wrote:
> On 07/11/2011 02:04 AM, H.J. Lu wrote:
>>
>> With my original change,  I got
>>
>> (const:DI (plus:DI (symbol_ref:DI ("iplane.1577") [flags 0x2]
>> <var_decl 0x7ffff0857960 iplane>)
>>         (const_int -4 [0xfffffffffffffffc])))
>>
>> I think it is safe to permute the conversion and addition operation
>> if one operand is a constant and we are zero-extending.  This is
>> how zero-extending works.
>
> Ok, I think I understand what you mean.  The key is the
>
>   XEXP (x, 1) == convert_memory_address_addr_space
>                  (to_mode, XEXP (x, 1), as)
>
> test.  It ensures basically that the constant has 31-bit precision, because
> otherwise the constant would change from e.g. (const_int -0x7ffffffc) to
> (const_int 0x80000004) when zero-extending it from SImode to DImode.
>
> But I'm not sure it's safe.  You have,
>
>  (zero_extend:DI (plus:SI FOO:SI) (const_int Y))
>
> and you want to convert it to
>
>  (plus:DI FOO:DI (zero_extend:DI (const_int Y)))
>
> (where the zero_extend is folded).  Ignore that FOO is a SYMBOL_REF (this
> piece of code does not assume anything about its shape); if FOO ==
> 0xfffffffc and Y = 8, the result will be respectively 0x4 (valid) and
> 0x100000004 (invalid).

This example contradicts what you said above "It ensures basically that the
constant has 31-bit precision".  For zero-extend, the issue is address-wrap.
As I understand, to support address-wrap, you need to use ptr_mode.

> If pointers extend as signed you also have a similar case.   If FOO ==
> 0x7ffffffc and Y = 8, the result of
>
>  (sign_extend:DI (plus:SI FOO:SI) (const_int Y))
>
> and
>
>  (plus:DI FOO:DI (sign_extend:DI (const_int Y)))
>
> will be respectively 0xffffffff80000004 (valid) and 0x80000004 (invalid).
>
>
> What happens if you just return NULL instead of the assertion (good idea
> adding it!)?
>
> Of course then you need to:
>
> 1) check the return values of convert_memory_address_addr_space_1, and
> propagate NULL up to simplify_unary_operation;
>
> 2) check in simplify-rtx.c whether the return value of
> convert_memory_address_1 is NULL, and only return if the return value is not
> NULL.  This is not yet necessary (convert_memory_address is the last
> transformation for both SIGN_EXTEND and ZERO_EXTEND) but it is better to
> keep code clean.

I will give it a try.

Thanks.


-- 
H.J.

  reply	other threads:[~2011-07-11 15:54 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-07-05 14:30 H.J. Lu
2011-07-09 21:22 ` H.J. Lu
2011-07-09 21:23 ` Paolo Bonzini
2011-07-09 21:41   ` H.J. Lu
2011-07-10 17:04     ` Paolo Bonzini
2011-07-10 21:16       ` H.J. Lu
2011-07-11  0:48         ` H.J. Lu
2011-07-11  1:14           ` H.J. Lu
2011-07-11  6:49             ` H.J. Lu
2011-07-11 11:09             ` Paolo Bonzini
2011-07-11 15:58               ` H.J. Lu [this message]
2011-07-11 16:57                 ` H.J. Lu
2011-07-11 17:26                   ` H.J. Lu
2011-07-13 16:24                 ` Paolo Bonzini
2011-07-13 16:52                   ` H.J. Lu
2011-07-13 16:55                     ` Paolo Bonzini
2011-07-13 16:58                       ` Paolo Bonzini
2011-07-13 18:42                         ` H.J. Lu
2011-07-25 10:34                           ` Paolo Bonzini
2011-07-27 18:18                             ` H.J. Lu
2011-07-27 22:41                               ` Paolo Bonzini
2011-07-28  3:11                                 ` H.J. Lu
2011-07-28  7:59                                   ` Paolo Bonzini
2014-05-29  4:52                     ` Andrew Pinski
2014-05-29 16:13                       ` H.J. Lu
2014-05-29 16:23                         ` pinskia
2014-05-29 17:09                           ` H.J. Lu
2014-05-29 17:20                             ` pinskia
2014-05-30  7:18                               ` Paolo Bonzini
2011-07-28 10:24 Uros Bizjak
2011-07-28 10:27 ` Paolo Bonzini
2011-07-29 13:29   ` H.J. Lu
2011-07-28 18:23     ` Uros Bizjak
2011-07-29  9:32     ` Paolo Bonzini
2011-07-29  9:37       ` Paolo Bonzini
2011-07-30  0:47       ` H.J. Lu
2011-07-30 16:35         ` Paolo Bonzini

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='CAMe9rOp7=UvknX2z9k5O1Q8sJkXH8Z11=7TEjU0M5_EoAFC11g@mail.gmail.com' \
    --to=hjl.tools@gmail.com \
    --cc=bonzini@gnu.org \
    --cc=gcc-patches@gcc.gnu.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 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 read-only IMAP folder(s) and NNTP newsgroup(s).