qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Richard Henderson <rth@twiddle.net>
To: Peter Maydell <peter.maydell@linaro.org>
Cc: QEMU Developers <qemu-devel@nongnu.org>
Subject: Re: [Qemu-devel] [PATCH v2 41/60] target-i386: Create gen_lea_v_seg
Date: Fri, 27 Dec 2013 06:49:12 -0800	[thread overview]
Message-ID: <52BD9368.7090902@twiddle.net> (raw)
In-Reply-To: <CAFEAcA-1-UWD96noKYPE54anuP=RWr48yboZD48ApLa2O63H8A@mail.gmail.com>

On 12/26/2013 01:27 PM, Peter Maydell wrote:
>> The only time s->addseg will be false in 16-bit mode is during translation of
>> LEA.  I do need the addseg check there for LEA cleanup, but this change should
>> not affect gen_string_movl.
> 
> Oh, is that the bit that does:
> 
>         val = s->addseg;
>         s->addseg = 0;
>         gen_lea_modrm(env, s, modrm);
>         s->addseg = val;
> 
> ? I think we should get rid of that -- s->addseg should always
> mean "we can do the segment-base-is-zero optimization",
> it shouldn't be a secret hidden parameter to the gen_lea functions
> saying "don't do this addition".

Perhaps.  I'd rather do that as follow-up, since lea_v_seg isn't the top-level
function called for LEA.  And if we clean up addseg, we might as well clean up
popl_esp_hack as well.

How about a comment for now?

>> I disagree with the characterization "random temporary".  Using the output as a
>> temporary in computing the output is totally sensible, given that our most
>> popular host platform is 2-address.
> 
> This is the kind of thing that in an ideal world the register allocator
> would deal with. The tcg/README optimisation suggestions
> don't say anything about preferring to use X,X,Y rather than X,Y,Z
> ops where possible, and typically allocating and using a special
> purpose temp is more readable code.

I agree that a real register allocator would handle it.

I disagree that a special purpose temp would result in more readable code,
since then we'd need to add *more* code to deallocate it after the other
arithmetic.

> More generally, it's pretty unclear to me why we're handling
> "use default segment register for instruction" (ie ovr_seg == -1)
> differently for the three cases.

Because the handling of segments is fundamentally different in each of the
three cpu modes?

> Why is it OK to skip the addition of the base address for ES
> (in the movl_A0_EDI case) when the comment for addseg says
> it only applies to CS/DS/ES?

Err... when are we skipping the addition in gen_string_movl_A0_EDI?  We pass in
R_ES as the segment register to use...

> It seems to me that we ought to try to get this code to a
> point where it looks more like:
>     if (ovr_seg < 0) {
>         ovr_seg = def_seg;
>     }
>     emit code to get address;
>     if (!segment_base_guaranteed_zero(s, ovr_seg)) {
>         emit code to add base to address;
>     }
> 
> where segment_base_guaranteed_zero() is a helper
> function like:
> bool segment_base_guaranteed_zero(s, seg) {
>       /* Return true if we can guarantee at translate time that
>        * the base address of the specified segment is zero
>        * (and thus can skip emitting code to add it)
>        */
>       return (!s->addseg &&
>           (seg == R_CS || seg == R_DS || seg == R_SS));
> }

That does look much nicer, I agree.


r~

  parent reply	other threads:[~2013-12-27 14:49 UTC|newest]

Thread overview: 75+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-11-29  2:59 [Qemu-devel] [PATCH v2 00/60] target-i386 improvements Richard Henderson
2013-11-29  2:59 ` [Qemu-devel] [PATCH v2 01/60] exec: Delay CPU_LOG_TB_CPU until we actually execute a TB Richard Henderson
2013-11-29  2:59 ` [Qemu-devel] [PATCH v2 02/60] target-i386: Push DisasContext into load/store helpers Richard Henderson
2013-11-29  2:59 ` [Qemu-devel] [PATCH v2 03/60] target-i386: Stop encoding DisasContext.mem_index Richard Henderson
2013-11-29  2:59 ` [Qemu-devel] [PATCH v2 04/60] target-i386: Use new tcg_gen_qemu_ld_* helpers Richard Henderson
2013-11-29  2:59 ` [Qemu-devel] [PATCH v2 05/60] target-i386: Use new tcg_gen_qemu_st_* helpers Richard Henderson
2013-11-29  2:59 ` [Qemu-devel] [PATCH v2 06/60] target-i386: Replace OT_* constants with MO_* constants Richard Henderson
2013-11-29  2:59 ` [Qemu-devel] [PATCH v2 07/60] target-i386: Remove gen_op_ld_T0_A0 Richard Henderson
2013-11-29  2:59 ` [Qemu-devel] [PATCH v2 08/60] target-i386: Remove gen_op_ldu_T0_A0 Richard Henderson
2013-11-29  2:59 ` [Qemu-devel] [PATCH v2 09/60] target-i386: Remove gen_op_ld_T1_A0 Richard Henderson
2013-11-29  2:59 ` [Qemu-devel] [PATCH v2 10/60] target-i386: Remove gen_op_lds_T0_A0 Richard Henderson
2013-11-29  2:59 ` [Qemu-devel] [PATCH v2 11/60] target-i386: Introduce gen_op_st_rm_T0_A0 Richard Henderson
2013-11-29  2:59 ` [Qemu-devel] [PATCH v2 12/60] target-i386: Remove gen_op_st_T0_A0 Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 13/60] target-i386: Remove gen_op_st_T1_A0 Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 14/60] target-i386: Fix typo in gen_push_T1 Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 15/60] target-i386: Tidy mov[sz][bw] Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 16/60] target-i386: Tidy movsl Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 17/60] target-i386: Remove unused arguments to gen_lea_modrm Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 18/60] target-i386: Use MO_BE for movbe Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 19/60] target-i386: Tidy gen_op_mov_TN_reg+tcg_gen_trunc_tl_i32 Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 20/60] target-i386: Tidy load + truncate Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 21/60] target-i386: Tidy extend + store Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 22/60] target-i386: Tidy extend + move Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 23/60] target-i386: Remove gen_op_movl_T0_0 Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 24/60] target-i386: Remove gen_op_movl_T0_im* Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 25/60] " Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 26/60] target-i386: Remove gen_op_mov*_A0_im Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 27/60] target-i386: Remove gen_movtl_T*_im Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 28/60] target-i386: Remove gen_op_andl_T0_ffff Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 29/60] target-i386: Remove gen_op_andl_T0_im Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 30/60] target-i386: Remove gen_op_movl_T0_T1 Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 31/60] target-i386: Remove gen_op_andl_A0_ffff Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 32/60] target-i386: Use TCGMemOp for 'ot' variables Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 33/60] target-i386: Change gen_op_add_reg_* size parameter to TCGMemOp Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 34/60] target-i386: Change gen_op_j*z_ecx " Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 35/60] target-i386: Change aflag " Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 36/60] target-i386: Change gen_op_mov_reg_A0 size parameter " Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 37/60] target-i386: Change dflag " Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 38/60] target-i386: Tidy addr16 code in gen_lea_modrm Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 39/60] target-i386: Combine gen_push_T* into gen_push_v Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 40/60] target_i386: Clean up gen_pop_T0 Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 41/60] target-i386: Create gen_lea_v_seg Richard Henderson
2013-12-26 18:38   ` Peter Maydell
2013-12-26 19:31     ` Richard Henderson
2013-12-26 21:27       ` Peter Maydell
2013-12-26 21:31         ` Peter Maydell
2013-12-27 14:49         ` Richard Henderson [this message]
2013-12-27 16:06           ` Peter Maydell
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 42/60] target-i386: Use gen_lea_v_seg in gen_lea_modrm Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 43/60] target-i386: Use gen_lea_v_seg in stack subroutines Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 44/60] target-i386: Tidy cpu_regs initialization Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 45/60] target-i386: Access segs via TCG registers Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 46/60] target-i386: Use gen_lea_v_seg in pusha/popa Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 47/60] target-i386: Rewrite gen_enter inline Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 48/60] target-i386: Introduce mo_stacksize Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 49/60] target-i386: Rewrite leave Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 50/60] target-i386: Remove gen_op_mov_reg_T0 Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 51/60] target-i386: Remove gen_op_mov_reg_T1 Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 52/60] target-i386: Remove gen_op_addl_T0_T1 Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 53/60] target-i386: Remove gen_op_mov_TN_reg Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 54/60] target-i386: Remove gen_op_mov_reg_A0 Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 55/60] target-i386: Remove gen_op_movl_A0_reg Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 56/60] target-i386: Tidy gen_add_A0_im Richard Henderson
2013-12-26 18:58   ` Peter Maydell
2013-12-26 19:10     ` Richard Henderson
2013-12-26 22:34       ` Peter Maydell
2013-12-27 15:17         ` Richard Henderson
2013-12-27 15:32           ` Peter Maydell
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 57/60] target-i386: Tidy some size computation Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 58/60] target-i386: Rename gen_op_jmp_T0 to gen_op_jmp_v Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 59/60] target-i386: Tidy ljmp Richard Henderson
2013-11-29  3:00 ` [Qemu-devel] [PATCH v2 60/60] target-i386: Deconstruct the cpu_T array Richard Henderson
2013-12-23 20:15 ` [Qemu-devel] [PATCH v2 00/60] target-i386 improvements Richard Henderson
2013-12-23 22:54   ` Peter Maydell
2013-12-26 19:03     ` Peter Maydell

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=52BD9368.7090902@twiddle.net \
    --to=rth@twiddle.net \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-devel@nongnu.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 NNTP newsgroup(s).