qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] cvttps2dq, movdq2q, movq2dq incorrect behaviour


From: malc
Subject: Re: [Qemu-devel] cvttps2dq, movdq2q, movq2dq incorrect behaviour
Date: Wed, 21 Jun 2006 12:21:07 +0400 (MSD)

On Wed, 21 Jun 2006, Julian Seward wrote:


Malc, your sse-movq.patch works for me.  Thanks.

soft-float was a red herring, translate.c is at fault here (interpreter
does not use it, hence behaved correctly)

translate.c:3009
if (b1 >= 2 && ((b >= 0x50 && b <= 0x5f) ||
                 b == 0xc2)) {
     /* specific case for SSE single instructions */
     if (b1 == 2) {
         /* 32 bit access */
         gen_op_ld_T0_A0[OT_LONG + s->mem_index]();
         gen_op_movl_env_T0(offsetof(CPUX86State,xmm_t0.XMM_L(0)));
     } else {
         /* 64 bit access */
         gen_ldq_env_A0[s->mem_index >>
2](offsetof(CPUX86State,xmm_t0.XMM_D(0))); }
} else {
     gen_ldo_env_A0[s->mem_index >> 2](op2_offset);
}

cvttps2dq is 0x5b(b=0x5b) with repn prefix (b1=2) the above code is
optimized a bit more than it should have been, as it loads only 4 bytes
into xmm_t0 instead of 16.

Uh, fine, but I don't understand how/what to fix.  Can you advise?

Following will fix the _specific_ case of cvttps2dq, ideally one
should go through all the [0x50..0x5f, 0xc2] with (repnz,repz prefix)
range and check wether the rules imposed by the above snippet apply.

--- /mnt/big/npf/cvs/qemux/qemu/target-i386/translate.c Tue Jun 20 15:19:20 2006
+++ translate.c Tue Jun 20 18:17:19 2006
@@ -3009,7 +3009,9 @@
                 if (b1 >= 2 && ((b >= 0x50 && b <= 0x5f) ||
                                 b == 0xc2)) {
                     /* specific case for SSE single instructions */
-                    if (b1 == 2) {
+                    if (b1 == 2 && b == 0x5b) {
+                        gen_ldo_env_A0[s->mem_index >> 
2](offsetof(CPUX86State,xmm_t0.XMM_Q(0)));
+                    } else if (b1 == 2) {
                         /* 32 bit access */
                         gen_op_ld_T0_A0[OT_LONG + s->mem_index]();
                         
gen_op_movl_env_T0(offsetof(CPUX86State,xmm_t0.XMM_L(0)));

--
mailto:address@hidden




reply via email to

[Prev in Thread] Current Thread [Next in Thread]