From: malc <malc@pulsesoft.com>
To: qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] cvttps2dq, movdq2q, movq2dq incorrect behaviour
Date: Tue, 20 Jun 2006 15:29:10 +0400 (MSD) [thread overview]
Message-ID: <Pine.LNX.4.64.0606201524080.2085@home.oyster.ru> (raw)
In-Reply-To: <200606201154.40985.jseward@acm.org>
[-- Attachment #1: Type: TEXT/PLAIN, Size: 1313 bytes --]
On Tue, 20 Jun 2006, Julian Seward wrote:
>
> The SSE2 instructions cvttps2dq, movdq2q, movq2dq do not behave
> correctly, as shown by the attached program. It should print
>
> cvttps2dq_1 ... ok
> cvttps2dq_2 ... ok
> movdq2q_1 ... ok
> movq2dq_1 ... ok
>
> but instead produces
>
> cvttps2dq_1 ... ok
> cvttps2dq_2 ... not ok
> result0.sd[0] = 12 (expected 12)
> result0.sd[1] = 3 (expected 56)
> result0.sd[2] = -2147483648 (expected 43)
> result0.sd[3] = 3 (expected 87)
> movdq2q_1 ... not ok
> result0.uq[0] = 1302123111658042420 (expected 5124095577148911)
> movq2dq_1 ... not ok
> result0.uq[0] = 1302123111658042420 (expected 5124095577148911)
> result0.uq[1] = 6221254864647256184 (expected 0)
>
> I looked at QEMU's instruction decoders for these, and compared them
> to Valgrind's, but could not see what the problem was. The decode
> logic looks OK. Maybe the problem is elsewhere.
The signature of movdq2q is Pq, VRq and for movq2dq - Vo, PRq it appears
that translate.c gets it backwards, attached patch should deal with it.
As for cvttps2dq i ran it with interpreter which uses outdated(i.e. non
soft-float) conversion routines and it passed, so my guess would be that
this is float32_to_int32_round_to_zero vs (int32_t) cast issue.
--
mailto:malc@pulsesoft.com
[-- Attachment #2: Type: TEXT/PLAIN, Size: 1181 bytes --]
--- translate.c Tue Jun 20 15:19:01 2006
+++ /mnt/big/npf/cvs/qemux/qemu/target-i386/translate.c Tue Jun 20 15:19:20 2006
@@ -2947,15 +2947,15 @@
case 0x2d6: /* movq2dq */
gen_op_enter_mmx();
rm = (modrm & 7) | REX_B(s);
- gen_op_movq(offsetof(CPUX86State,xmm_regs[reg & 7].XMM_Q(0)),
- offsetof(CPUX86State,fpregs[rm].mmx));
- gen_op_movq_env_0(offsetof(CPUX86State,xmm_regs[reg & 7].XMM_Q(1)));
+ gen_op_movq(offsetof(CPUX86State,xmm_regs[rm].XMM_Q(0)),
+ offsetof(CPUX86State,fpregs[reg & 7].mmx));
+ gen_op_movq_env_0(offsetof(CPUX86State,xmm_regs[rm].XMM_Q(1)));
break;
case 0x3d6: /* movdq2q */
gen_op_enter_mmx();
rm = (modrm & 7);
- gen_op_movq(offsetof(CPUX86State,fpregs[reg].mmx),
- offsetof(CPUX86State,xmm_regs[rm].XMM_Q(0)));
+ gen_op_movq(offsetof(CPUX86State,fpregs[rm].mmx),
+ offsetof(CPUX86State,xmm_regs[reg].XMM_Q(0)));
break;
case 0xd7: /* pmovmskb */
case 0x1d7:
next prev parent reply other threads:[~2006-06-20 11:30 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-06-20 10:54 [Qemu-devel] cvttps2dq, movdq2q, movq2dq incorrect behaviour Julian Seward
2006-06-20 11:29 ` malc [this message]
2006-06-20 11:48 ` Julian Seward
2006-06-20 14:26 ` malc
2006-06-21 0:31 ` Julian Seward
2006-06-21 8:21 ` malc
2006-06-21 11:04 ` malc
2006-06-21 23:01 ` Julian Seward
2006-06-20 13:15 ` RE : " Sylvain Petreolle
2006-06-20 13:51 ` malc
2006-06-20 14:13 ` Julian Seward
2006-06-20 15:06 ` RE : " Sylvain Petreolle
2006-06-20 15:14 ` Guillaume POIRIER
2006-06-20 14:17 ` RE : " Jens Axboe
2006-06-20 14:23 ` 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=Pine.LNX.4.64.0606201524080.2085@home.oyster.ru \
--to=malc@pulsesoft.com \
--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).